informit.com -- Your Brain is Hungry.

InformIT - Introduction From: Using Oracle8

You are here : Home : Using Oracle8

Introduction
Exact Phrase All Words Search Tips

Contents

Next >

Save to MyInformIT

From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q q q q q q q

Introduction Who Should Use This Book Why This Book? How This Book Is Organized Conventions Used in This Book About the Authors Acknowledgments Tell Us What You Think!

Introduction
Welcome to Using Oracle8! This book identifies the many functions an Oracle DBA needs to perform on an Oracle8 database and explains how to do them as efficiently and effectively as possible. You learn about the key functions of database administration, including installing the product, designing and creating a database and its tablespaces, designing and creating the tables and other objects that make up an Oracle database, designing and executing a good backup strategy with a recovery methodology, and monitoring and tuning performance. You also learn about creating and maintaining users and performing an upgrade to Oracle8, as well as other tasks that you may need in your position as DBA. You also learn when and how to use the various tools Oracle8 provides to assist you in database management, performance monitoring and tuning, data loading, backup and recovery, and data export and import. The book is designed to let you read about a topic at length when you have the time and the inclination, or to use as a quick reference guide when you need an answer to a pressing technical question or an example to follow when performing a specific task. Using Oracle8 contains cross-references to related topics so that you can look at all aspects of a topic, even if they're covered in different chapters. These cross-references also enable you to read the book in any order you choose. If you run across a subject you don't fully understand, you can easily switch your attention to the area(s) identified and carry on your reading there. Where applicable, the book also references the Oracle documentation materials, so you can find even more detail if you need it. Don't forget to keep this book handy at work, just in case you need to check something in a hurry that you haven't read about yet or is a new topic to you. Be sure also to use the tear-out card inside the book's cover. It contains some of the most common, but difficult to remember, information you'll need.

Who Should Use This Book

You'll get the most out of this book if you have some background in the SQL language and some knowledge or experience with relational databases. Because Oracle's SQL language is based on the ANSI standard, it's not discussed in detail in this book, but numerous examples use SQL statements. The theory of relational databases is also outside the scope of this book, as are the internal structures within Oracle, except where they're needed to

http://www.informit.com/content/0789716534/element_000.shtml (1 of 6) [26.05.2000 16:45:57]

informit.com -- Your Brain is Hungry. InformIT - Introduction From: Using Oracle8

help you understand how or why to perform a specific task. This book is intended primarily for DBAs who have some knowledge of relational databases. Much of this book will be familiar if you've worked with earlier releases of Oracle-but you'll find the new Oracle8 features discussed. If you've worked with other relational databases, you may need to refer to the glossary if you find brand new terms or terms that have different meanings in Oracle. If you haven't worked with any relational databases, you should expect to follow the frequent cross-references to sections of the book; this will fill in background information as you read about a topic.

Why This Book?

Have you ever purchased a Using book from Que? The Using books have proven invaluable to readers as both learning guides and as references for many years. The Using series is an industry leader and has practically become an industry standard. We encourage and receive feedback from readers all the time, and consider and implement their suggestions whenever possible. This book isn't a compiled authority on all the features of Oracle8; instead, it's a streamlined, conversational approach for using Oracle8 productively and efficiently. This book has many features: q Improved index. What do you call tasks and features? As we wrote this book, we anticipated every possible name or description of a feature or database activity. q Real-life answers. Throughout the book you find our real-life examples and experiences. We recommend how to organize your database on the logical as well as the physical level. We suggest what values to use when assigning physical storage attributes to your tables, indexes, and other database objects, and how to determine if you have made a good set of choices. After all, we've been there, done that! We understand that how to perform a task is only one question you may have, and perhaps the bigger questions are "Why?" and "What for?" q Relevant information written just for you. We have carefully scrutinized which features and tasks to include in this book and have included those that apply to your everyday use of Oracle. Why invest in material that teaches you how to perform tasks you may never need to perform? q Reference or tutorial. You can learn to quickly perform a task using step-by-step instructions, or you can investigate the why and wherefore of a task with our discussions preceding each task. q Wise investment. We don't waste your valuable bookshelf space with redundant or irrelevant material, nor do we assume you "know it all" or need to know it all. Here is what you need, when you need it, how you need it, with an appropriate price tag. q Easy-to-find procedures. Every numbered step-by-step procedure in the book has a short title explaining exactly what it does. This saves you time by making it easier to find the exact steps you need to accomplish a task.

How This Book Is Organized

Using Oracle8 has task-oriented, easy-to-navigate tutorials and reference information presented in a logical progression from simple to complex tasks. It covers features of the program you use in your daily work. Examples are real life. You can work through the book lesson by lesson, or you can find specific information when you need to perform a job quickly. Using Oracle8 is divided into nine parts: q Part I, Building Your Oracle Database. Part I introduces you to relational databases in general and to the basic tools used to build and manage an Oracle8 database, whether you're creating a database from scratch or converting from an earlier release. q Part II, Customizing Your Oracle Database. Part II shows you how to build the appropriate storage units for your database objects. It's here you also find out how to manage the shared structures required for Oracle to function in the multiuser environment, including redo log files, rollback segments, and temporary segments. q Part III, Managing Data and Index Segments. Part III provides the details needed to build Oracle database tables and indexes, including information on sizing them and assigning appropriate physical characteristics and alternate structures. You find out about the logical and physical design of data and index information, including optional segment structures such as index-organized tables. q Part IV, Managing Users and Database Security. Part IV explains how to create user ids and manage user access to the database and its objects. You are also introduced to methods for monitoring and controlling resource usage. The chapters in this section include detailed information on the new password-management features introduced in Oracle8.
http://www.informit.com/content/0789716534/element_000.shtml (2 of 6) [26.05.2000 16:45:57]

informit.com -- Your Brain is Hungry. InformIT - Introduction From: Using Oracle8

q

Part V, Backing Up Your Oracle Database. Part V covers the various options available for protecting your database contents against loss due to hardware failure. You learn how to restore data that's lost when failures occur. The chapters in this section also cover the Recovery Manager tools, introduced as part of Oracle8. Part VI, Tuning Your Database Applications. In Part VI you learn about the various tools and techniques that DBAs and application developers should consider when building and tuning applications. These include performance-analysis tools and various resource-saving design considerations such as indexes, clustering techniques, optimizer selection, and constraint management. Part VII, Tuning Your Database. Part VII addresses the issues related to gathering and analyzing performance information about your database. The chapters in this section include information on tools available for these tasks, as well as how to interpret various statistics available to you and how to respond to performance degradation caused by various factors. Part VIII, Using Additional Oracle Tools and Options. In Part VIII you learn about the various tools provided by Oracle as part of the base product that can help you manage your database and the data within it, plus network access between your applications and the database. This section also summarizes the features available with the products that you can optionally license for added functionality if needed, such as Oracle Parallel Server and the Object option. Additional information available at our Web site (www.mcp.com/info). Appendix A, "Essential PL/SQL: Understanding Stored Procedures, Triggers, and Packages," includes a comprehensive guide to the PL/SQL language and the database constructs you can build with it. Appendix B, "What's New to Oracle8," lists the Oracle8's new features for those of you who are familiar with earlier Oracle releases and just need to identify what changes you may want to study and implement in your own database. Appendix C, "Installing Oracle8," covers the basic steps you need to follow in order to install a new version of the database, whether you're upgrading from Oracle7 or installing Oracle8 directly.

Now look at the detailed table of contents, decide what you want to read now or in the near future, and begin getting comfortable with Oracle8.

Conventions Used in This Book

The following items are some of the features that will make this book easier for you to use: q Cross-references. We've looked for all the tasks and topics related to a topic at hand and referenced them for you. If you need to look for coverage that leads up to what you're working on, or if you want to build on the new skill you just mastered, you have the references to easily find the right coverage in the book: SEE ALSO Information on tablespace usage for different segment types, see page xxx q Glossary terms. For all terms that appear in the glossary, you'll find the first appearance of that term italicized in the text. q SideNotes. Information related to the task at hand or "inside" information from the authors is offset in SideNotes, so they won't interfere with the flow of the text, yet make it easy to find valuable information. Each SideNote has a short title to help you quickly identify the information you'll find there. Oracle's syntax for commands, scripts, and SQL statements also incorporate special elements. Look at the following syntax example: ALTER DATABASE [database_name] ADD LOGFILE [GROUP [group_number]] filename [SIZE size_integer [K|M]] [REUSE] Terms that are italicized are considered placeholders. When you use the command, you replace the italicized word with an appropriate value. For example, database_name in the preceding code would be replaced with an actual database name. Square brackets ([]) in command syntax indicate optional clauses. The brackets around [database_name] in the preceding code indicate that you aren't required to provide a database name. Don't include the brackets when you use the command. The | character indicates that you choose one item or the other, not both. (For example, you can choose either K for kilobytes or M for megabytes for the preceding command.) Again, don't use this character in the actual command. Ellipses (...) in listings indicate either a clause that can repeat or skipped code that's not pertinent to the discussion. Don't use the ellipses in the actual code.

http://www.informit.com/content/0789716534/element_000.shtml (3 of 6) [26.05.2000 16:45:57]

informit.com -- Your Brain is Hungry. InformIT - Introduction From: Using Oracle8

q

Line numbers are included in some code listings to make discussion about the code easier to reference. Don't include the numbers with any command-line commands, as part of any Oracle scripts, or within SQL statements.

About the Authors

David Austin has been in the data processing profession for almost 25 years. He had worked with many database architectures, including hierarchical, net-work, and relational, before becoming an Oracle DBA about 10 years ago. For the past five years, David has worked for Oracle Corporation, where he's now employed as a senior principal curriculum developer. His previous positions at Oracle include senior principal consultant and senior principal instructor. David has completed multiple Masters programs with Oracle Education and is a Certified Oracle Database Administrator. David is a contributing author for Que's Special Edition Using Oracle8 and wrote Chapters 3, 5-8, 10-11, 16-17, and the glossary for this book. He obtained both of his degrees-a B.A. with a double major in mathematics and English and a minor in computer science and an M.S. in applied mathematics-from the University of Colorado. Vijay Lunawat is a technical specialist with Oracle Corporation; he resides in Orlando, Florida. He has a bachelor's degree in electronics engineering. He has worked with Oracle databases for more than 10 years as a developer, database administrator, consultant, and support. A specialist in Oracle Parallel Server, he's now working with the Center of Expertise in Oracle Support Services. He develops and frequently teaches Oracle Internals classes at Oracle. He was a contributing author for Que's Special Edition Using Oracle8 and wrote Chapters 1, 13, 18, 27, and Appendix A for this book. Meghraj Thakkar works as a senior technical analyst at Oracle Corporation. He has been working with various Oracle products for the past six years. He has a master's in computer science and a bachelor's in electronics engineering. He has several industry vendor certifications, including Microsoft Certified Systems Engineer (MCSE), Novell Certified ECNE, and Lotus Certified Notes Consultant. He has taught several courses at the University of California, Irvine; developed and presented a two-day course, "Supporting Oracle on Windows NT," to internal Oracle employees; and presented two papers at the ECO'98 held in New York City in March 1998. He also coauthored several books for Macmillan Computer Publishing- Special Edition Using Oracle8, Oracle8 Server Unleashed, and Oracle8 for DBAs-and contributed Chapters 2, 4, 14, and 21-23 for this book. Tomas Gasper , the author of Chapters 9, 12, and 20, is an Oracle DBA for Energizer Battery Company in St. Louis, Missouri. He has worked in a variety of system support roles, including DBA, UNIX, and Windows NT administrator and systems programmer. As a refugee from the defense industry, Tomas enjoys learning about exotic and unique computer systems. His hobbies include experimenting with Linux systems, Web-based applications, and, of course, exploring the Internet. Tomas can be reached at [email protected] or [email protected]. Ari Kaplan, the author of Chapter 25, is an independent computer consultant specializing in Oracle product design, development, and management. Ari, coauthor of Que's Special Edition Using Oracle8 and Waite Group Press's Oracle8 How-To, both for Macmillan Computer Publishing, has played pivotal roles in implementing some of the nation's largest and most visible Oracle applications for various industries. Ari worked for Oracle before becoming a consultant in 1994, and since 1990 he has worked as a consultant for several Major League Baseball clubs (currently for the Montreal Expos), where he has developed and managed their scouting department's software systems. Ari graduated from the California Institute of Technology in Pasadena and was granted the school's prestigious "Alumni of the Decade" distinction for his contributions in the computer industry. He has appeared on NBC's Today Show and CNN, and is a frequent guest speaker on Oracle in the United States and abroad. Ari lives in Chicago and can be reached at [email protected] or through his Web site of Oracle tips at http://homepage.interaccess.com/~akaplan. Raman Batra, the author of Chapters 24 and 26, is a database administrator for the Cessna Aircraft Company in Wichita, Kansas, where he has been working since 1994. In the past, Raman has developed client/server and intranet applications using Developer/2000 Forms, Pro*C, and Oracle Web Application Server. As a DBA, he is administering Designer/2000 applications and application suites for engineering, supply chain, and parts distribution systems on Oracle7 and 8. He is also designing an advanced replication architecture on Oracle8 for failover and disaster recovery. Raman is a member of the Oracle Technology Network, ACM, IOUGA, and the Wichita Area Oracle Users Group. Raman can be reached at [email protected]. Joseph Duer, the author of Chapters 15 and 19, is a technical analyst and Oracle database administrator at a technology-driven corporation based in southern Connecticut. He specializes in Web development using Oracle Database Server and Oracle Application Server. He has developed object-oriented systems that utilize C++,

http://www.informit.com/content/0789716534/element_000.shtml (4 of 6) [26.05.2000 16:45:57]

informit.com -- Your Brain is Hungry. InformIT - Introduction From: Using Oracle8

Java, and JavaScript, as well as Oracle Application Server's Java, PL/SQL, and VRML cartridges. He can be reached via email at [email protected] and via his Web page at http://www.netcom.com/~joeduer.

Acknowledgments
From David Austin: Thanks to the many professionals who have helped and encouraged me in my career and my work on this book, particularly my various managers, instructors, and colleagues at Oracle, including Deborah West, Chris Pirie, Nick Evans, Vijay Venkatachalam, Larry Mix, Beth Winslow, Sue Jang, Scott Gossett, and Scott Heisey. I would also like to say thank you to some of my earliest mentors in this business-Bob Klein, Donald Miklich, and Roland Sweet, wherever they might be. Thanks also to the various editors at Que who helped shepherd this work from its inception to the book you now have in your hands, with a special mention for Angela Kozlowski and Susan Dunn. I also want to thank my coauthors, without whose efforts this work could never have been finished. Finally, a thank you to my family for putting up with the long hours I spent ignoring them while working on this project. My wife, Lillian, is now bracing for the revisions, while my kitten is just happy that she once again gets some petting when she sits in my lap. From Vijay Lunawat: Most thanks go to my two children, Siddharth and Sanchi, and my wife, Sushma, for their patience and for putting up with my long and weekend working hours while I was writing for this book. From Meghraj Thakkar: I would like to give a special thanks to my wife, Komal, for her patience and understanding. From Raman Batra: To my lovely wife, Sarika, for her understanding and admirable patience in keeping my daughter, Nikita, away from me, when I was writing. Nikita had a real hard time understanding why Daddy was working with such "boring" text stuff with no music, when she could be watching her Winnie the Pooh CD on Daddy's PC. From Joe Duer: I would like to thank once again the Tuesday night crew at Shelton EMS: Jason, Betty, John, and Denise. Your help covering all the shifts I missed because I was writing is greatly appreciated. I would like to thank everyone at Que-in particular, Angela Kozlowski and Susan Dunn-for their help and guidance during the development of this book.

Tell Us What You Think!

As the reader of this book, you are our most important critic and commentator. We value your opinion and want to know what we're doing right, what we could do better, what areas you'd like to see us publish in, and any other words of wisdom you're willing to pass our way. As the Executive Editor for the Client/Server Database Team at Macmillan Computer Publishing, I welcome your comments. You can fax, email, or write me directly to let me know what you did or didn't like about this book-as well as what we can do to make our books stronger. Please note that I cannot help you with technical problems related to the topic of this book, and that due to the high volume of mail I receive, I might not be able to reply to every message. When you write, please be sure to include this book's title and author as well as your name and phone or fax number. I will carefully review your comments and share them with the author and editors who worked on the book. Fax: E-mail: Mail: 317-817-7070 [email protected] Executive Editor Client/Server Database Team Macmillan Computer Publishing 201 West 103rd Street Indianapolis, IN 46290 USA

http://www.informit.com/content/0789716534/element_000.shtml (5 of 6) [26.05.2000 16:45:57]

informit.com -- Your Brain is Hungry. InformIT - Introduction From: Using Oracle8

Contents

Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_000.shtml (6 of 6) [26.05.2000 16:45:57]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Introducing Relational Databases and Oracle8

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q

What's a Database Management System? Oracle Database Files

r r r r r

The Initialization Parameter File The Control File The Data File Redo Log Files Archived Redo Log Files Starting and Stopping Instances Oracle Enterprise Manager (OEM) SQL*Plus PL/SQL Net8 Precompilers Developer/2000 Statistics and the Data Dictionary Dynamic Performance Tables

Understanding Database Instances

r

Oracle8's Tools
r r r r r r

The Oracle8 Data Dictionary

r r

q q q q

Functions performed by a database management system Physical architecture of an Oracle database Identify the major components of an Oracle instance Overview of database tools: Oracle Enterprise Manager, SQL*Plus, PL/SQL, Net8, Developer 2000, and precompilers Understand the Oracle8 data dictionary and dynamic performance views

What's a Database Management System?

A database can be defined as a collection of information organized in such a way that it can be retrieved and used. A database management system (DBMS) can further be defined as the tool that enables us to manage and
http://www.informit.com/content/0789716534/element_001.shtml (1 of 12) [26.05.2000 16:46:05]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8

interact with the database. Most DBMSs perform the following functions: q Store data q Create and maintain data structures q Allow concurrent access to many users q Enforce security and privacy q Allow extraction and manipulation of stored data q Enable data entry and data loading q Provide an efficient indexing mechanism for fast extraction of selected data q Provide consistency among different records q Protect stored data from loss by backup and recovery process Several different types of DBMSs have been developed to support these requirements. These systems can broadly be classified in the following classes: q A hierarchical DBMS stores data in a tree-like structure. It assumes a parent-child relationship between the data. The top of the tree, known as the root, can have any number of dependents. Dependents, in turn, can have any number of subdependents, and so on. Hierarchical database systems are now obsolete. q A network DBMS stores data in the form of records and links. This system allows more flexible many-to-many relationship than do hierarchical DBMSs. Network DBMSs are very fast and storage-efficient. Network database management systems allowed complex data structures but were very inflexible and required tedious design. An airline reservation system is one example of this type of DBMS system. q Relational DBMSs (RDBMSs) probably have the simplest structure a database can have. In an RDBMS, data is organized in tables. Tables, in turn, consist of records, and records of fields. Each field corresponds to one data item. Two or more tables can be linked (joined) if they have one or more fields in common. RDBMSs are easy to use and have flourished in the last decade. They're commonly used on low-end computer systems. In the last few years, however, their use has expanded to more powerful computer systems. Oracle, Informix, and Sybase are some popular RDBMSs available in the market. Oracle8 stores objects in relational tables Oracle8 is an object relational database management system, which allows objects to be stored in tables, in a manner similar to numbers and words being stored in an RDBMS system.
q

Object-oriented DBMSs were designed to handle data such as numbers and words. During recent years, however, object-oriented DBMSs are emerging. These systems can handle objects such as videos, images, pictures, and so on.

Oracle Database Files

An Oracle database physically resides in various files. Figure 1.1 shows the physical structure of an Oracle database. Figure 1.1 : An Oracle database system.

The Initialization Parameter File

The parameter file, commonly known as INIT.ORA, contains initialization parameters that control the behavior and characteristics of the database and the instance that accesses the database. You can edit this text file in your favorite editor. Changing the parameter file The initialization parameter file is read by the instance only during startup. Any changes made in the initialization file take effect only after you shut down and restart the instance. Oracle supplies a sample INIT.ORA file in the $ORACLE_HOME/dbs directory. $ORACLE_HOME is the top-level directory under which Oracle software is installed; it doesn't need to be the user Oracle's home directory. The default name of an instance's parameter file is initSID.ora, in which SID (System IDentifier) is a character string that uniquely identifies the instance on the system. You can override the defaults by using the PFILE parameter of the Server Manager's startup command. The
http://www.informit.com/content/0789716534/element_001.shtml (2 of 12) [26.05.2000 16:46:05]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8

IFILE parameter in this file allows you to nest multiple initialization files for the same instance.

The Control File

The control file contains information about the database's physical structure and status. It has information about several things: the total number of data files; log files; redo log groups; redo log members; current redo log to which the database is writing; name and location of each data file and online redo log files; archived log history; and so on. Starting from Oracle8, the control file also contains information about the backup of the database. Oracle updates the control file Oracle automatically records any structural changes in the database- for example, addition/deletion of a data file-in the control file(s). An Oracle instance updates the control file(s) with various status information during its operation. The control_file initialization parameter specifies the name and location of a database's control file. It's strongly recommended that you specify multiple files in the control_file initialization parameter to mirror the control file to multiple locations. The V$CONTROL_FILE data dictionary view contains information about the database's control file. The V$CONTROLFILE_RECORD_SECTION dynamic performance view contains detailed structure information about the control file. This view gives the information about all records contained in the control file.

The Data File

An Oracle database stores user information in physical data files. A data file can contain tables, indexes, clusters, sequences, data dictionary, rollback segments, temporary segments, and so on. At the logical level, Oracle manages space in terms of tablespace (a group of one or more data files). When an Oracle database is created, it has only one tablespace: SYSTEM. Other tablespaces and the associated data files are added later, as needed. You can specify the name, location, and size of a data file while creating the tablespace to which the data file belongs. Oracle uses control files to store the name and location of the data files. Use the data dictionary views V$DATAFILE and DBA_DATA_FILES to retrieve the information about a database's data files.

Redo Log Files

Oracle records all changes against the database in the redo log file and uses the contents of the redo log file to regenerate the transaction changes in case of failure. An Oracle database has two or more redo log files. Oracle allows you to mirror the redo log files, thus a redo log group contains one or more files (members). Oracle writes to all the members of a redo log group simultaneously. An Oracle instance writes to redo log groups in cyclical order-that is, it writes to one redo log group and then to the next when the earlier one is filled up. When the last available redo log group is filled, it switches over to the first one. SEE ALSO For detailed information about redo log files, You can specify the name, location, and size of the redo log files during database creation. The V$LOGFILE data dictionary view contains redo log files' information. You can also add, delete, and relocate redo log files by using the ALTER DATABASE command.

Archived Redo Log Files

The archived log file contains a copy of the redo log file. Archived redo log files are useful in recovering the database and all committed transactions in case of failures (such as disk failure). When an Oracle database is operating in archive log mode, it needs to archive the recently filled redo log file before it can reuse it. You can enable automatic archiving by setting the initialization parameter LOG_ARCHIVE_START to TRUE or by issuing the archive log start command after the instance startup. When automatic archiving is enabled, the ARCH (archiver) process copies the filled redo log files to the directory specified by LOG_ARCHIVE_DEST. The LOG_ARCHIVE_FORMAT parameter defines the default names of archived log files.

Understanding Database Instances

An Oracle database stores data in physical data files and allows controlled user-access to these files through a

http://www.informit.com/content/0789716534/element_001.shtml (3 of 12) [26.05.2000 16:46:05]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8

set of operating system processes. These processes are started during the instance startup. Because they work silently, without direct user interaction, they're known as background processes. To enable efficient data manipulation and communication among the various processes, Oracle uses shared memory, known as Shared Global Area (SGA). These background processes and the shared memory segment together are referred as an Oracle instance. In a parallel server environment, a database can be accessed by multiple instances running on different machines. An Oracle instance consists of the following background processes: q The process monitor process (PMON). This background process cleans up after a user process terminates abnormally. It rolls back the uncommitted transaction left behind and releases the resources locked by the process that no longer exists. Oracle background processes that are always started The LMON, PMON, DBWR, and LGWR processes are always present for an instance. Other processes are started by setting up a related initialization parameter.
q

The database writer process (DBWR). To ensure efficient and concurrent data manipulation, Oracle doesn't allow a user process to directly modify a data block on the disk. The blocks that need to be modified or in which the data is inserted are first fetched in a common pool of buffers, known as buffer cache. These blocks are then written to the disk in batches by the DBWR background process. Thus, DBWR is the only process with write access to the Oracle data files. The log writer process (LGWR). Whenever an Oracle process modifies an Oracle data block, it also writes these changes to the redo log buffers. It's the responsibility of the LGWR process to write the redo log buffers to the online redo log file. This process reads the contents of the redo log buffers in batches and writes them to the online redo log file in sequential fashion. Note that LGWR is the only process writing to the online redo log files. Oracle's transaction commit algorithm ensures that the contents of redo log buffers are flushed to the online redo log file whenever a transaction is committed. The system monitor process (SMON). This background process does operations such as freeing up the sort space and coalescing the adjacent free extents in one big extent. SMON is also responsible for performing transaction recovery during the instance recovery (during instance startup after a crash or shutdown abort). In a parallel server environment, it also detects and performs instance recovery for another failed instance. The archiver process (ARCH). This process is started when the database is in archive log mode and automatic archiving is enabled. It copies the recently filled online redo log file to an assigned backup destination. The checkpoint process (CKPT). During a checkpoint, the DBWR process writes all the modified blocks to disk. DBWR also tells LGWR to update the header information of all data files with the checkpoint information. Because a database containing a larger amount of data files might be a time-consuming task for LGWR, the CKPT process starts during instance startup to help LGWR update the file headers during the checkpoint. This process is started only when the CHECKPOINT_PROCESS parameter is set to TRUE or when the number of data files in the database is more than a certain number. The recoverer process (RECO). This process is responsible for recovering the in-doubt transaction in a distributed database environment. This process is started only when the initialization parameter DISTRIBUTED_TRANSACTION is set to greater than 0. The parallel query slave processes (pxxx). Under favorable conditions, Oracle can reduce the execution time for certain SQL operations by dividing the operation among several dedicated processes. The processes used for parallel execution of SQL statements are known as parallel query slaves. The snapshot process (SNPn). The snapshot or the job queue processes are started when the parameter JOB_QUEUE_PROCESSES is set more than 0. These processes execute jobs in the job queue, refresh any snapshot that's configured for automatic refresh, and so on. The dispatcher process (Dxxx). Oracle supports multithreaded servers on some operating systems. When enabled, these processes receive the user request and put it in the request queues for execution. They also collect the results of the execution from the dispatcher queues and pass them back to users. The shared server process (Sxxx). In a multithreaded server environment, the dedicated server process executes the SQL operations from the request queues and puts back the results in the corresponding dispatcher queue.

If you're running Oracle's parallel server option, you also see the following background processes on each instance: q The lock process (Lckn). This Oracle parallel server process coordinates all the lock requests from the local and remote instances. It communicates with user processes and the lock daemon process.

http://www.informit.com/content/0789716534/element_001.shtml (4 of 12) [26.05.2000 16:46:05]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8
q

The lock monitor process (LMON). This Oracle parallel server process is responsible for reconfiguring the Integrated Distributed Lock Manager (IDLM) during instance startup and shutdown in an OPS environment. It also performs lock cleanup after abnormal death of a user process. The lock daemon process (LMD0). This Oracle parallel server process is part of the IDLM. It handles all lock requests from remote instances for the locks held by a local instance.

Figure 1.2 shows the components of an Oracle instance: the SGA and background processes. Figure 1.2 : An Oracle instance consists of the SGA and background processes.

Starting and Stopping Instances

An Oracle database isn't accessible to users until it's opened by an Oracle instance. In an Oracle parallel server environment, an Oracle database is accessed by more than one instance. Each instance has its own set of background processes and the SGA. An instance startup operation involves starting all the background processes and allocating the shared memory area (see Figure 1.3). Who can start up and shut down an Oracle instance? An instance startup operation can be done only by users with the requisite OS privileges or who have been assigned an OSOPER or OSDBA role. Figure 1.3 : Oracle instance startup consists of three steps. How Oracle starts instances 1. Oracle starts all the background processes and allocates the SGA. Oracle reads the initialization parameter file during this step. 2. Oracle reads the control file and associates the control with the instances. It detects the conditions of the database from the last shutdown/crash. 3. Oracle reads the file headers of all the data files and redo log files. It ensures consistency among all data files. If the instance is being started after a crash or shutdown abort, Oracle also applies all the redo log since the last successful checkpoint. Oracle database is accessible to users after completing this step. If the instance is started after a crash or shutdown abort, Oracle needs to perform rollback operations for the uncommitted transaction. This operation is performed by SMON in the background while the database is open and available for use. An Oracle instance shutdown closes the database, dismounts it, and then removes the SGA and the background processes. Shutdown offers three modes: normal, immediate, and abort. Shutdown normal and shutdown immediate are used most often, whereas shutdown abort should be used with caution. During shutdown normal, Oracle waits for all users to disconnect, writes all modified data to the data files, and then updates files headers, online redo log files, and control files. Shutdown immediate disconnects all users and then proceeds similarly to shutdown immediate. Shutdown abort just removes all the background processes and the SGA; all cleanup work is done during the next startup. Table 1.1 lists Server Manager commands to start and stop an Oracle instance. Table 1.1 Server Manager startup and shutdown commands Command: startup and startup pfile=file startup nomount startup mount alter database mount alter database open shutdown shutdown immediate Description: Uses the default parameter file to start startup open the instance, mount the database, and open it Starts the instance by using the specified parameter file Allocates the SGA and starts the background processes; doesn't mount and open the database Allocates the SGA, starts the background process, and mounts the database; doesn't open the database Mounts the database after the instance is started with startup nomount command Opens the database after it's mounted by the startup mount command Closes the instance after all users disconnect (normal shutdown) Doesn't allow any new transactions to start; rolls back uncommitted transactions and closes the instance

http://www.informit.com/content/0789716534/element_001.shtml (5 of 12) [26.05.2000 16:46:05]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8

shutdown abort

Immediately removes the SGA and the background processes

SEE ALSO To learn how to start and stop database instances with Oracle Enterprise Manager,

Oracle8's Tools
Oracle provides various tools for application development and for performing administrative functions: q Oracle Enterprise Manager (OEM) q SQL*Plus q PL/SQL q Net8 q Developer 2000 q Precompilers

Oracle Enterprise Manager (OEM)

Oracle Enterprise Manager is a graphical system management tool that allows you to perform multiple tasks in a complicated database environment. OEM comes with several components. Some components, such as Oracle Expert and Performance Manager, are priced separately. Chapter 4 "Managing with Oracle Enterprise Manager (OEM)," explains how to use these components. OEM's major components are as follows (see Figure 1.4): Figure 1.4 : Oracle Enterprise Manager consists of several modules. Backup Manager Security Manager Data Manager Storage Manager Instance Manager Schema Manager SQL Worksheet
q

Software Manager Oracle Expert Lock Manager TopSession Monitor Performance Manager Tablespace Manager

Backup Manager lets you perform backup and recovery operations associated with the database. It lets you interface with Oracle8's advanced backup and recovery utility, the Recovery Manager. Data Manager, a data transfer and loading tool, lets you invoke export, import, and load utilities. Use the Export utility to extract data in Oracle's operating system-independent format. The exported data can be loaded in another Oracle database or later in the same database. You also can use Export as the database's logical backup. The Loader utility is used to insert data in the Oracle database from text files. Instance Manager lets you manage instances, user sessions, and in-doubt transactions. It lets you start and shut down Oracle instances. You can manage multiple instances in an Oracle parallel server environment. It also lets you manage the initialization parameter file used during instance startup. Lock Manager lets you view the locks held in an instance. It's a helpful tool for analyzing hung sessions and other, similar situations. Oracle Expert lets you tune instance and database performance. It generates a listing of recommendations that can be implemented automatically to improve the performance. Performance Manager lets you monitor an Oracle instance performance. It provides you with graphical representation of various performance statistics. Schema Manager lets you perform Data Definition Language (DDL) operations, which let you create, alter, drop, and view database objects such as tables, indexes, clusters, triggers, and sequences. Security Manager lets you perform user-management tasks such as adding, altering, and dropping users, roles, and profiles. Software Manager allows you to administer in an distributed environment and to automate database administration tasks. SQL Worksheet behaves mostly similar to a SQL*Plus session. You can use it to enter and execute SQL

http://www.informit.com/content/0789716534/element_001.shtml (6 of 12) [26.05.2000 16:46:05]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8

commands. Storage Manager lets you perform database space-management tasks such as creating, altering, and dropping tablespaces. It also lets you create online/offline and drop rollback segments. Tablespace Manager lets you view the space usage within a tablespace at object level. You can also get information about used and free space within the database. TopSession Monitor lets you monitor active user sessions and view user resource utilization. This information can be used to address slow performance. Oracle Trace lets you drill down the execution of SQL statements to improve performance of the system.

SQL*Plus
The only interface available between end users and an RDBMS is Structured Query Language (SQL). All other applications and tools that users utilize to interact with the RDBMS act as translators/interpreters. These tools generate SQL commands based on a user's request and pass the generated SQL commands on to the RDBMS. SQL*Plus can't start or stop an instance A database administrator can't start and shut down an Oracle instance by using SQL*Plus. SQL*Plus, Oracle's version of SQL, is one of the most commonly used Oracle tools. SQL*Plus enables users to instruct the Oracle instance to perform the following SQL functions: q Data definition or DDL operations, such as creating, altering, and dropping database objects q Data query to select or retrieve the stored data q Data manipulation or the DML operations to insert, update, and delete data q Access and transfer data between the databases q Allow user to enter data interactively q DBA functions or the database administrative tasks such as managing users (creating, altering, and dropping users), managing space (creating, altering, and dropping tablespaces), and backup and recovery In addition to these basic SQL functions, SQL*Plus also provides several editing and formatting functions that enable users to print query results in report format. Setting Up the SQL*Plus Environment SQL*Plus has many advanced functions that you can use to present data in a visually pleasing format. You can set various environment variables in order to control the way SQL*Plus outputs a query. Table 1.2 lists some of the most common commands to set up the environment, which you can enter at the SQLPLUS> prompt. Table 1.2 SQL*Plus environment commands Command: set set set set set set pagesize linesize newpage pause array feedback Description: Sets the number of lines per page Sets the number of characters in a line Sets the number of blank lines between pages Causes SQL*Plus to pause before each page Sets the number of rows retrieved at a time Displays the number of records processed by a query Prints a heading at the beginning of the report Allows output from DBMS_OUTPUT.PUT_LINE stored procedure to be displayed Displays timing statistics Allows you to suppress output generated by a command executed from a file

set heading set serveroutput set time set term

Set up the environment automatically You also can use the LOGIN.SQL and GLOGIN.SQL files to set up the environment for the current session while invoking SQL*Plus.

http://www.informit.com/content/0789716534/element_001.shtml (7 of 12) [26.05.2000 16:46:05]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8

PL/SQL
PL/SQL stands for Procedural Language/Structured Query Language. It allows a user to utilize structured programming constructs similar to third-generation languages such as C, Fortran, and COBOL. PL/SQL enhances SQL by adding the following capabilities: PL/SQL is embedded in Oracle8 tools Although you can use PL/SQL as a programming language, it's also available as part of Oracle tools such as Oracle Forms and Oracle Reports. The PL/SQL engine embedded in these tools acts as the preprocessor.
q q q q q q

Define and use variables Control the program flow (IF, IF...THEN...ELSE, and FOR LOOP constructs) Use of cursors and arrays File I/O Functions and procedures PL/SQL tables to move larger amount of data

With PL/SQL, you can use SQL commands to manipulate data in an Oracle database and also use structured programming constructs to process the data.

Net8
Net8, formerly known as SQL*Net, is Oracle's networking interface. It allows communication between various Oracle products residing on different machines. It enables communication among client, server, and Oracle databases in a distributed environment. At the client end, the client application code passes messages on to the Net8 residing locally, and the local Net8 transfers messages to the remote Net8 via the underlying transport protocol. These messages are received by Net8 at the server, which sends them to the database server for execution. The server executes the request and responds to the client following the same path. Figure 1.5 shows the communication between client and server using Net8. Figure 1.5 : The client and the server communicate with each other through Net8. Net8 has many enhancements over its predecessor SQL*Net, such as connection pooling, multiplexing, listener load balancing, and caching the network addresses at the client end. Net8 is backward-compatible and can coexist with SQL*Net version 2.

Precompilers
A third-generation language compiler doesn't recognize the SQL needed to interface with the RDBMS. Therefore, if you need power and flexibility of a language such as C, C++, Fortran, or COBOL and also want it to interface with the Oracle8 RDBMS, you need a tool that can convert the SQL statements to the calls that a language compiler can understand. As Figure 1.6 shows, a precompiler program reads structured source code and generates a source file that a language compiler can process. Oracle provides several precompilers, such as Pro*C, Pro*Cobol, Pro*Fortran, and Pro*Pascal. Figure 1.6 : You develop programs by using a precompiler. You might want to use precompilers to get better performance while developing long-running batch programs and time-critical programs. You can do the following by using precompilers: q Use dynamic SQL. q Better control cursor and program flow. q Develop program libraries and use them in multiple applications. q Concurrently access the data from multiple databases. q Write multithreaded applications by forking processes.

Developer/2000
Developer/2000 provides the complete set of tools to develop applications that access an Oracle database. It consists of tools for creating forms, reports, charts, queries, and procedures. It also enables you to deploy existing and new applications on the Web. Developer/2000 consists of the following component tools: q Project Builder. You can track and control application documents, source code files, charts, forms,

http://www.informit.com/content/0789716534/element_001.shtml (8 of 12) [26.05.2000 16:46:05]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8

reports, queries, and so on. Forms Builder. Forms are one of the easiest and most popular means for end users to interact with the database. End users totally unfamiliar with the database and the SQL language can easily learn a forms-based application to access the Oracle database. Forms Builder is the tool application developers can use to develop forms. Report Builder. Although forms give users online/immediate interaction with the database, the Report Builder allows users to queue the data extraction request in a queue at a remote report server. The report server, in turn, interacts with the database to generate the reports. You can also embed the reports in other online tools such as Web browsers, graphics, and forms. Graphics Builder. Graphical and visual representation of the data is much more effective than raw data. You can use Graphics Builder to produce interactive graphical displays. Graphics Builder also allows you to include graphs in forms and reports. Query Builder. Query Builder allows you to interact with database tables in tabular form onscreen. The tables involved in the query are available onscreen, and users can construct the desired query by pointing and clicking. Formatted query results are also displayed. Schema Builder. Schema Builder is a graphical DDL (data definition language) tool. You can use it to create, alter, and drop database objects such as tables, indexes, clusters, and sequences. Procedure Builder. The Procedure Builder helps you build procedures interactively. You can use its graphical interface to create, edit, debug, and compile PL/SQL programs. The PL/SQL programs unit, which can be generated with the Procedure Builder, includes packages, triggers, functions, and program libraries. Translation Builder. The Translation Builder allows you to extract translatable strings from Oracle or non-Oracle resources and perform the desired transaction. For example, you can translate Microsoft Windows (.RC) and HTML files to Oracle resource files.

Traditionally, Developer/2000 supported the client/server architecture, where the client tools and the application reside on one machine (usually the end-user PC) and the database server resides on another machine. With the proliferation of the Web, however, Oracle has introduced a three-tier ar rchitecture in which an additional server that runs the application code has been introduced. Client/server, or the three-tier, architecture for installing Developer/2000 is highly recommended because the workload is distributed among the client, database server, and application servers in this structure. In addition, the application, Developer/2000, and the database software are independent of each other, thus making maintenance easier. SQL*Net or Net8 needs to be installed on the client and the database server to enable the connectivity between the two.

The Oracle8 Data Dictionary

Oracle stores information about all the objects defined by the users, structural information about the database, and so on in its internal tables. These Oracle internal tables and associated objects are collectively referred as the data dictionary. The data dictionary is owned by the user SYS and always resides in the SYSTEM tablespace. Data dictionary tables are created when the database is created Oracle automatically updates these tables whenever it needs to. Users should never update any table in the data dictionary. Several Oracle and non-Oracle tools also create some objects in the data dictionary that are used for storing operational, reference, and configuration information. Information stored in the data dictionary is available to users through data dictionary views. A database administrator or a user can use the data dictionary to view the following information: q Definitions of the database objects such as tables, partitions, indexes, clusters, views, snapshots, triggers, packages, procedures, functions, sequences, and synonyms q Users defined in the database q Storage allocation for the objects in the database and quota assigned to each user q Integrity constraints q Database links q Privileges and roles q Replicated objects, snapshots, and their refresh characteristics q Auditing information such as access patterns for various objects
http://www.informit.com/content/0789716534/element_001.shtml (9 of 12) [26.05.2000 16:46:05]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8
q q q q q q

Jobs in the job queue Locks and latches held in the database Alerts and table queues (advance queues) Rollback segments SQL*Loader direct load NLS settings

Oracle's data dictionary views can broadly be defined in the following classes: q Views with the DBA prefix. These views contain information about the entire database. For example, the view DBA_TABLES gives information about all tables in the database. By default, these views are accessible only to users with the DBA role. q Views with the USER prefix. USER views contain information about the objects owned by the user. For example, USER_TABLES gives information about the tables owned by the user. q Views with the ALL prefix. These views contain information about all objects accessible to the user. Objects accessible to a user include the objects created by the user plus the objects on which he has received grants from other users. For example, the ALL_TABLES view contains information about all tables accessible to a user. Table 1.3 lists important Oracle8 data dictionary views. Similar views with DBA and ALL prefixes are available. Table 1.3 Important data dictionary views View Name: USER_ALL_TABLES USER_CLUSTERS USER_CONSTRAINTS USER_DB_LINKS USER_ERRORS USER_EXTENTS USER_FREE_SPACE USER_INDEXES USER_IND_COLUMNS USER_JOBS USER_RESOURCE_LIMITS USER_SEGMENTS USER_SEQUENCES USER_SNAPSHOTS USER_SYNONYMS USER_TAB_COLUMNS USER_TAB_PARTITIONS USER_TABLES USER_TRIGGERS Description: Contains descriptions of all tables available to the user Contains information about clusters created by the user Contains information about the constraint defined by the user Contains information about the database link created by the user Gives all current errors on all stored objects for the user Lists all the extents used by the objects owned by the user Lists all free extents in the tablespaces on which the user has privilege Gives information about indexes created by the user Gives the name of all the columns on which the user has created indexes Gives all jobs in the job queue owned by the user Gives resource limits applicable for the user Gives information about all segments owned by the user Lists information about all sequences owned by the user Gives information about all snapshots the user can view Gives the name of all private synonyms for the user Gives the name of all columns in all tables the user owns Gives information about all table partitions owned by the user Gives information about all tables the user owns Gives information for all triggers created by the user

Statistics and the Data Dictionary

Several data dictionary views contain columns with statistics information for the object. For example, the USER_TABLES view contains columns NUM_ROWS (number of rows in the table), BLOCKS (number of data blocks used in the table), AVG_ROW_LEN (average row length of a row in the table), and so on. These columns are populated only when you analyze the object by using the ANALYZE command. You should analyze the objects at regular intervals to keep the statistics up-to-date.

Dynamic Performance Tables

An Oracle instance maintains comprehensive information about its current configuration and activity. These

http://www.informit.com/content/0789716534/element_001.shtml (10 of 12) [26.05.2000 16:46:05]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8

statistics are accessible to the database administrator through dynamic performance views. Most of these views are based on in-memory table-like structures known as virtual tables (because they aren't real tables). The majority of these views have names starting with V$. These virtual tables don't require disk storage space and aren't stored in any tablespace. By default, the dynamic performance views are accessible to the SYS user or to the users having a SYSDBA role. Contents of these views are updated continuously while the instance is active. Use TIMED_STATISTICS to gather timing information Many dynamic performance views contain columns, such as WAIT_TIME and TOTAL_WAITS, that contain timing information. Such columns are populated by Oracle only when the TIMED_STATISTICS parameter is set to TRUE. Table 1.4 describes important dynamic performance views. These views are for Oracle8; some may not exist in Oracle7. Table 1.4 Dynamic performance views View Name: V$ACCESS V$CONTROLFILE V$DATABASE V$DATAFILE V$DATAFILE_HEADER V$DB_LINK V$FILESTAT V$FIXED_TABLE Description: Displays information about locked database objects and the sessions accessing them Lists names of the database control files Contains miscellaneous database information such as database name creation date, archive/no archive log mode, and so on Contains information about the data files that are part of the database (This information is from the control file.) Similar to V$DATAFILE, except that information is based on the contents of each data file header Lists information about all active database links Displays read/write statistics for each database data file Contains names of all fixed tables in the database

V$FIXED_VIEW_DEFINITION Lists definitions of all the dynamic performance views; you can see how Oracle creates dynamic performance views based on its internal x$ tables; these x$ tables are known as fixed tables Lists license-related information V$LICENSE Shows the locks held and requested; information in this view useful V$LOCK while tuning the database performance or hanging issues Lists all the objects locked in the database and the sessions that are V$LOCKED_OBJECT locking the objects Lists information about the online redo logs V$LOG Contains information about the archived redo log file V$LOG_HISTORY Lists statistics about the current session V$MYSTAT Lists current values of the initialization parameters; the V$PARAMETER ISDEFAULT column indicates whether the parameter value is the default V$PROCESS Lists all Oracle processes; a value of 1 in the BACKGROUND column indicates that the process is an Oracle background process; a NULL value in this column indicates a normal user process Used to query the information about the files needing media recovery; this view can be queried after the instance mounts the database Lists names of all the online rollback segments Lists statistics for all online rollback segments Contains information about all the current sessions; this view, one of the most informative, has about 35 columns Contains information about waits each session has incurred on events; use this view if you're experiencing slow performance

V$RECOVER_FILE

V$ROLLNAME V$ROLLSTAT V$SESSION V$SESSION_EVENT

http://www.informit.com/content/0789716534/element_001.shtml (11 of 12) [26.05.2000 16:46:06]

informit.com -- Your Brain is Hungry. InformIT - Introducing Relational Databases and Oracle8 From: Using Oracle8

V$SESSION_WAIT V$SESSTAT V$SESS_IO V$STATNAME V$SYSSTAT V$SYSTEM_EVENT V$TABLESPACE V$TRANSACTION V$WAITSTAT

Lists the events and resources Oracle is waiting on; information in this view can be used to detect performance bottlenecks Contains performance statistics for each active session Lists I/O statistics about each active session Gives names of Oracle statistics displayed in V$SESSTAT and V$SYSSTAT Contains performance statistics for the whole instance Contains information for various Oracle events Lists names of all tablespaces in the database Lists statistics related to transactions in the instance Contains block contention statistics

Global dynamic performance views In a parallel server environment, every V$ view has a corresponding GV$ view. These views, known as global dynamic performance views, contain information about all active instances of an Oracle parallel server environment. The INST_ID column displays the instance number to which the information displayed in the GV$ view belongs. Use fixed tables with caution! Oracle doesn't encourage the use of fixed tables listed in V$FIXED_TABLE because their structure isn't published and can be changed. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_001.shtml (12 of 12) [26.05.2000 16:46:06]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

You are here : Home : Using Oracle8

Creating a Database
Exact Phrase All Words Search Tips

< Back

Contents

Next >

Save to MyInformIT

From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q q

Prerequisites for Creating a Database Choosing Initialization Parameters for your Database Getting Ready to Create a Database
r r r r

Organizing the Database Contents Designing a Database Structure to Reduce Contention and Fragmentation Decide on the Database Character Set Start the Instance Using the Oracle Installer (ORAINST) to Create a Database Using the CREATE DATABASE Command

Choosing the Method for Creating the Database

r r

q q

Creating a Database from the Seed Database Checking the Status of your Database
r r

Examining Basic Views Checking the Oracle Alert Log

q q q q q

Create a new database Create an Oracle service for Windows NT Run optional data dictionary scripts Understand the initialization parameters Use the alert log

Prerequisites for Creating a Database

Before you can create an Oracle database, you need to configure the kernel for shared memory. In UNIX, shmmax needs to be set properly to allow for the total SGA set in the initialization file. In Windows NT, make sure that virtual memory isn't more than twice the physical memory on the system. You also need to make certain decisions regarding how the database will be used and configured. These decisions should include the following: q Database sizing. Proper sizing of the database files can help you choose your initialization parameters. It will result in an optimally tuned database to start with. q Changing passwords for SYS and SYSTEM. These passwords will be used to log onto the database and perform administrative tasks. This can be done after database creation. The passwords of the SYS and the SYSTEM users should be changed as soon as possible. Use the ALTER USER command to change these passwords. It's recommended that you create a new account with SYSDBA privileges and use that account to create and own the database. Protect the password for SYS

http://www.informit.com/content/0789716534/element_002.shtml (1 of 13) [26.05.2000 16:46:14]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

Because SYS is the owner of the data dictionary, you should protect that password. Allowing the pass-word for SYS to get into the wrong hands can lead to tremendous damage to the database to the point that all data can be lost. The default password for SYS is CHANGE_ ON_INSTALL, whereas the default password for SYSTEM is MANAGER.
q

MTS configuration. The multithreaded server (MTS) pools connections and doesn't allocate a single thread per connection. As a result, it avoids stack overflow and memory allocation errors that would occur from a dedicated connection per thread. A multithreaded server configuration allows many user threads to share very few server threads. The user threads connect to a dispatcher process, which routes client requests to the next available server thread, thereby supporting more users. (You can configure MTS after database creation.) Setting the environment variables. Setting the ORACLE_SID, ORACLE_HOME, and path variables to the correct values will allow you to start the correct instance.

Prepare the operating system environment for database creation The operating system memory parameters should also be set properly. Check your operating system-specific documentation for the parameters to set. Prepare for and create a database (general steps) 1. Create the initSID.ora parameter file. 2. Create the configSID.ora file. 3. Create the database script crdbSID.ora. 4. Create the database. 5. Add rollback segments. 6. Create database objects for tools. The DBA's operating system privileges Your database administrator login should have administrator privileges on the operating system to be able to create a data-base.

Choosing Initialization Parameters for your Database

The instance for the Oracle database is started by using a parameter file (initSID.ora) that should be customized for the database. You can use the operating system to create this file by making a copy of the one provided by Oracle on the distribution media, or by using the init.ora file from the seed database (if installed) as a template. Rename this file as initSID.ora (for example, for the SID ABCD, the name of the initialization file would be initABCD.ora), and then edit it to customize it for your database. Change these parameters from their default values Most people make the mistake of leaving the initialization parameters to their default value. These default values aren't ideal for most systems. You need to carefully choose the initialization parameters with your data-base environment in mind. The parameter file is read-only at instance startup. If it's modified, you need to shut down and restart the instance for the new values to take effect. You can edit the parameter file with any operating system editor. Most parameters have a default value, but some parameters need to be modified with uniqueness and performance in mind. Table 2.1 lists parameters that should be specified. Table 2.1 Initialization parameters that you should modify Parameter: DB_NAME Description: Database identifier (maximum of eight characters). To change the name of an existing database, use the CREATE CONTROLFILE statement to recreate your control file(s) and specify a new database name. The network domain where the database is created. Names of the control files. If you don't change this parameter, the control files of other databases can be overwritten by the new instance, making the other instances unusable. Size in bytes of Oracle database blocks.

DB_DOMAIN CONTROL_FILES

DB_BLOCK_SIZE

http://www.informit.com/content/0789716534/element_002.shtml (2 of 13) [26.05.2000 16:46:14]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

SHARED_POOL_SIZE BACKGROUND_DUMP_DEST USER_DUMP_DEST DB_BLOCK_BUFFERS COMPATIBLE IFILE MAX_DUMP_FILE_SIZE PROCESSES ROLLBACK_SEGMENTS

Size in bytes of the shared pool. Location where background trace files will be placed. Location where user trace files will be placed. Number of buffers in the buffer cache. Version of the server that this instance is compatible with. Name of another parameter file included for startup. Maximum size in OS blocks of the trace files. Maximum number of OS processes that can simultaneously connect to this instance. Rollback segments allocated to this instance. Refer to the Oracle8 tuning manual for information and guidelines on determining the number and size of rollback segments based on the anticipated number of concurrent transactions. Number of bytes allocated to the redo log buffer in the SGA. Enable or disable automatic archiving if the database is in ARCHIVELOG mode. Default filename format used for archived logs. Location of archived redo log files. Maximum number of users created in the database.

LOG_BUFFER LOG_ARCHIVE_START LOG_ARCHIVE_FORMAT LOG_ARCHIVE_DEST LICENSE_MAX_USERS LICENSE_MAX_SESSIONS

Maximum number of concurrent sessions for the instance. LICENSE_SESSIONS_WARNING Warning limit on the concurrent sessions. Database names should be unique Attempting to mount two databases with the same name will give you the error ORA-01102: cannot mount database in EXCLUSIVE mode during the second mount. Setting the parameters The ideal values for these parameters are application dependent and are discussed in more detail in Chapter 21, "Identifying and Reducing Contention," and Chapter 22, "Tuning for Different Types of Applications." Setting these values is based on trial and error. For DSS systems, it's recommended that you choose a large value for these parameters; for OLTP systems, choose a small value for these parameters. The following is a sample init.ora file: db_name = SJR db_files = 1020 control_files = (E:\ORANT\database\ctl1SJR.ora, E:\ORANT\database\ctl2SJR.ora) db_file_multiblock_read_count = 16 db_block_buffers = 550 shared_pool_size = 9000000 log_checkpoint_interval = 8000 processes = 100 dml_locks = 200 log_buffer = 32768 sequence_cache_entries = 30 sequence_cache_hash_buckets = 23 #audit_trail = true #timed_statistics = true background_dump_dest = E:\ORANT\rdbms80\trace user_dump_dest = E:\ORANT\rdbms80\trace db_block_size = 2048 compatible = 8.0.3.0.0 sort_area_size = 65536 log_checkpoint_timeout = 0 remote_login_passwordfile = shared
http://www.informit.com/content/0789716534/element_002.shtml (3 of 13) [26.05.2000 16:46:14]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

max_dump_file_size = 10240 Create an initialization file 1. Copy the template file. In UNIX, copy $ORACLE_HOME/rdbms/install/rdbms/initx.orc to $ORACLE_HOME/dbs/initSID.ora. In Windows NT, copy $ORACLE_HOME\database\initorcl.ora to $ORACLE_HOME\database\initSID.ora. 2. Edit the initSID.ora by changing the following parameters: Parameter %pfile_dir% %config_ora_file% %rollback_segs% %init_ora_comments% UNIX Setting ?/dbs configSID.ora (created next) r01, r02, # Windows NT Setting ?/database configSID.ora (created next) r01, r02, #

Create configSID.ora 1. In UNIX, copy ?/rdbms/install/rdbms/cnfg.orc to ?/dbs/configSID.ora. In Windows NT, copy configorcl.ora to configSID.ora. 2. Edit the configSID.ora file with any ASCII text editor and set the following parameters: control_files, background_dump_dest, user_dump_dest, and db_name. Create the database script 1. Copy $ORACLE_HOME/rdbms/install/rdbms/crdb.orc to $ORACLE_HOME/dbs/crdbSID.sql. 2. Modify the crdbSID.sql file to set the following to the appropriate values: db_name, maxinstances, maxlogfiles, db_char_set, system_file, system_size, log1_file, log1_size, log2_file, log2_size, log3_file, and log3_size. When it's run, the crdbSID.sql does the following: q Runs the catalog.sql script, which will create the data dictionary q Creates an additional rollback segment, r0, in SYSTEM q Creates the tablespaces rbs, temporary, tools, and users q Creates additional rollback segments r01, r02, r03, and r04 in rbs q Drops the rollback segment r0 in SYSTEM q Changes temporary tablespaces for SYS and SYSTEM q Runs catdbsyn.sql as SYSTEM to create private synonyms for DBA-only dictionary views

Getting Ready to Create a Database

Creating a database is the first step in organizing and managing a database system. You can use the following guidelines for database creation on all operating systems. Check your operating system-specific documentation for platform-specific instructions. Before creating a database, take a complete backup of all your existing databases to protect against accidental modifications/deletions of existing files during database creation. The backup should contain parameter files, data files, redo log files, and control files. Mirror your control and redo log files The control files and redo log files help you recover your database. To keep from losing a control file, keep at least two copies of it active on different physical devices. Also, multiplex the redo log files and place the log group members on different disks. Also decide on a backup strategy and the size that will be required for online and archived redo logs. Backup strategies are discussed in Chapter 13, "Selecting and Implementing a Backup Strategy."

Organizing the Database Contents

You organize the database contents by using tablespaces. On some platforms, the Oracle installer creates a seed database, which has a number of predefined tablespaces. The tablespace structure should be carefully chosen by considering the characteristics of the data to minimize disk contention and fragmentation, and to improve

http://www.informit.com/content/0789716534/element_002.shtml (4 of 13) [26.05.2000 16:46:14]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

overall performance. In addition to the SYSTEM tablespace provided with the installation, Table 2.2 describes several other suggested tablespaces. You can create these tablespaces by using the CREATE TABLESPACE command, as shown later in the section "Using the CREATE DATABASE Command." Use multiple tablespaces Production data and indexes should be stored in separate tablespaces. Table 2.2 Suggested tablespaces to be created with the database Tablespace: TEMP RBS TOOLS APPS_DATA APPS_IDX Description: Used for sorting and contains temporary segments Stores additional rollback segments Tables needed by the Oracle Server tools Stores production data Store indexes associated with production data in APPS_DATA tablespace

Designing a Database Structure to Reduce Contention and Fragmentation

Separating groups of objects, such as tables with different fragmentation propensity, can minimize contention and fragmentation. You can use Table 2.3 as a guideline for separating objects. Table 2.3 Fragmentation propensity Segment Type: Data dictionary Rollback segments Temporary segments Application data Fragmentation: Zero Medium High Low

You can reduce disk contention by being familiar with the way in which data is accessed and by separating the data segments into groups based on their usage, such as separating q Segments with different backup needs q Segments with different security needs q Segments belonging to different projects q Large segments from smaller segments q Rollback segments from other segments q Temporary segments from other segments q Data segments from index segments Database sizing issues should be considered to estimate the size of the tables and indexes.

Decide on the Database Character Set

After the database is created, you can't change the character set without recreating the database. If users will access the database by using a different character set, the database character set should be the same as or a superset of all the character sets that would be used. Oracle8 uses encoding schemes that can be commonly characterized as single-byte 7-bit, single-byte 8-bit, varying-width multi-byte, and fixed-width multi-byte. Refer to the Oracle8 Server reference guide for limitations on using these schemes. SEE ALSO For more information on Oracle's National Language Support (NLS) feature and character sets,

Start the Instance

Make sure that the following parameters are set properly in the environment. If the following parameters aren't set properly, your instance won't start or the wrong instance might start: q ORACLE_SID. This parameter is used by Oracle to determine the instance to which the user will
http://www.informit.com/content/0789716534/element_002.shtml (5 of 13) [26.05.2000 16:46:14]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

q q q

connect. If ORACLE_SID isn't set properly and the CREATE DATABASE statement is run, you can wipe out your existing database and all its data. ORACLE_HOME. This parameter shows the full pathname of the Oracle system home directory. PATH. It should include $ORACLE_HOME. ORA_NLS. This is the path to the language object files. If ORA_NLS isn't set and the database is started with languages and character sets other than the database defaults, they won't be recognized.

After the following environment variables are verified, you can connect to Server Manager as internal and STARTUP NOMOUNT. Set the environment variables in UNIX 1. Set the ORACLE_SID variable as follows for the sh shell (XXX is your SID): ORACLE_SID XXX; export ORACLE_SID 2. Set the variable as follows for the csh shell: setenv ORACLE_SID XXX 3. Verify that ORACLE_SID has been set: Echo $ORACLE_SID 4. Start up the instance in nomount state: $svrmgrl SVRMGR> Connect internal SVRMGR> Startup nomount Set the environment variables in Windows NT 1. Use regedt32 to set the variables in the Registry's \HKEY_LOCAL_MACHINE\SOFTWARE\ORACLE hive. Or, from a DOS prompt, type C: > set ORACLE_SID=XXX where XXX is your SID name (maximum of four characters). 2. Use the Services tool in the Windows Control Panel to ensure that the ORACLESERVICESID service is started. Using Instance Manager on Windows NT On Windows NT, you can use the ORADIM utility (Instance Manager) to create a new instance and service for your database.

Choosing the Method for Creating the Database

You have several options to create the database: q You can use the Oracle installer to create a database-this is probably the easiest method of database creation because it allows the creation of a seed database, which you can use as a template to create new databases. Check your installation guide for platform-specific instructions for creating the seed database which has a fixed platform-specific schema. This method is discussed in the following section. q You can modify the create database scripts provided with Oracle as desired to create a database with your own schema. The name and location of these scripts varies with the operating system. On Windows 95/NT, the BUILDALL.SQL and BUILD_DB.SQL scripts can be used as a starting point for database creation. On UNIX systems, the crdbSID.sql is a similar script for database creation. By using this method, you can copy these mentioned scripts, make necessary changes to create the database, and then run the scripts to create the database. You can specify parameters such as MAXDATAFILES and specify multiple SYSTEM tablespace database files by using this method. q Manually create the database by executing the CREATE DATABASE command. Refer to the Oracle SQL reference manual for the complete syntax. This method allows for more flexibility by allowing you to specify MAXDATAFILES on parameters or multiple SYSTEM tablespace data files.But, there's also more possibility of syntax errors. After the database is created, you can run catalog.sql and catproc.sql while connected as the SYS of "internal"

http://www.informit.com/content/0789716534/element_002.shtml (6 of 13) [26.05.2000 16:46:14]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

account to create the data dictionary views. After the database is created, the SYSTEM tablespace and SYSTEM rollback segment will exist. A second rollback segment must be created and activated in the SYSTEM tablespace before any other tablespace can be created in the database. To create a rollback segment, from the Server Manager prompt type Svrmgr>Create rollback segment newsegment Tablespace system Storage (...); Refer to the SQL Language manual for the complete syntax of the CREATE ROLLBACK SEGMENT command.

Using the Oracle Installer (ORAINST) to Create a Database

This menu-driven method is probably the easiest because it runs the necessary scripts for any selected product. You can use this method to create a seed database. The installation guide for your platform should have specific instructions for this purpose. Oracle's installer isn't very flexible Using the Oracle installer for database creation isn't as flexible as the preceding methods in terms of specifying parameters such as MAXDATAFILES. If this method is used, you'll have to create the other standard non-system tablespaces.

Using the CREATE DATABASE Command

You also can create a database by using the SQL command CREATE DATABASE: CREATE DATABASE database [CONTROLFILE [REUSE]] [LOGFILE filespec[, ...]] MAXLOGFILES integer MAXLOGMEMBERS integer MAXLOGHISTORY integer DATAFILE filespec[, ...] MAXDATAFILES integer MAXINSTANCES integer ARCHIVELOG|NOARCHIVELOG EXCLUSIVE CHARACTERSET charset Table 2.4 lists the settings available with the CREATE DATABASE command. Table 2.4 CREATE DATABASE settings Option: database CONTROLFILE REUSE Description: The name of the database to be created. Specifies that existing control files specified by the CONTROL_FILES parameter can be reused. If REUSE is omitted and control files exist, you'll get an error. Specifies one or more files to be used as redo log files. Each filespec specifies a redo log file group containing one or more redo log file members or copies. If you omit this parameter, Oracle will create two redo log file groups by default. Specifies the maximum number of redo log file groups that can ever be created for this database. Specifies the maximum number of members or copies for a redo log file group. This parameter is useful only if you're using the PARALLEL SERVER option and in parallel and ARCHIVELOG mode. It specifies the maximum number of archived redo log files for automatic media recovery. Specifies one or more files to be used as data files.

LOGFILE

MAXLOGFILES MAXLOGMEMBERS MAXLOGHISTORY

DATAFILE

http://www.informit.com/content/0789716534/element_002.shtml (7 of 13) [26.05.2000 16:46:14]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

MAXDATAFILES MAXINSTANCES ARCHIVELOG or NOARCHIVELOG EXCLUSIVE CHARACTERSET

Specifies the maximum number of data files that can ever be created for this database. Specifies the maximum number of instances that can simultaneously have this parameter mounted and open. Establishes the mode for the redo log files groups. NOARCHIVELOG is the default mode. Mounts the database in the exclusive mode after it's created. In this mode, only one instance can access the database. Specifies the character set the database uses to store the data. This parameter can't be changed after the database is created. The supported character sets and default value of this parameter are operating system dependent.

Oracle performs the following operations when executing the CREATE DATABASE command: q Creates the data files as specified (if previously existing data files are specified, their data is erased) q Creates and initializes the specified control files q Creates and initializes the redo logs as specified q Creates the SYSTEM tablespace and the SYSTEM rollback segment q Creates the data dictionary q Creates the SYS and SYSTEM users q Specifies the character set for the database q Mounts and opens the database The data dictionary may not be created automatically You need to run the SQL scripts to create the data dictionary (catalog.sql and catproc.sql) if these scripts aren't run from your database creation script. The following example shows how to create a simple database: create database test controlfile reuse logfile GROUP 1 ('C:\ORANT\DATABASE\log1atest.ora', 'D:\log1btest.ora') size 500K reuse, GROUP 2 ( 'C:\ORANT\DATABASE\log2atest.ora', 'D:\log2btest.ora' ) size 500K reuse datafile 'C:\ORANT\DATABASE\sys1test.ora' size 10M reuse autoextend on next 10M maxsize 200M character set WE8ISO8859P1; This command creates a database called TEST with one data file (sys1test.ora) that's 10MB in size and multiplexed redo log files with a size of 500KB each. The character set will be WE8ISO8859P1.

Creating a Database from the Seed Database

The following steps can be used to create a database called MARS, using the starter (seed) database ORCL. If you don't have the starter database, you can use the sample initialization file INITORCL.80 in the c:\orant\database directory. Create a database in Windows NT with BUILD_ALL.sql 1. Create a directory called MARS. 2. Copy C:\ORANT\DATABASE\INITORCL.ORA to C:\MARS. 3. Modify the DB_NAME, CONTROL_FILES, GLOBAL_NAMES, and DB_FILES parameters in the INITMARS.ORA file. 4. Use the ORADIM80 command to create the service. For example, from a DOS prompt, type C: > oradim80 -NEW -SID TEST -INTPWD password

http://www.informit.com/content/0789716534/element_002.shtml (8 of 13) [26.05.2000 16:46:14]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

-STARTMODE AUTO -PFILE c:\orant\database\inittest.ora This command creates a new service called TEST, which is started automatically when Windows NT starts. INTPWD is the password for the "internal" account; the PFILE parameter provides the full pathname of initSID.ora. When to create Oracle services An Oracle serviceshould be created and started only if you want to create a database and don't have any other database on your system, or copy an existing database to a new database and retain the old data-base. 5. Set ORACLE_SID to MARS: C: > Set ORACLE_SID=MARS Copy the BUILD_DB.SQL script to c:\mars. Edit the BUILD_MARS.SQL script as follows: r Set PFILE to the full pathname for INITMARS.ORA. r Change CREATE DATABASE ORACLE to CREATE DATABASE MARS. r Change the data files and log filenames to the appropriate names. r Modify the location of the Oracle home directory to point to C:\MARS. Use Control Panel's Services tools to verify that the service ORACLESERVICEMARS is started. If it's not started, start it. Start Server Manager and connect to the database as "internal":

6. 7.

8. 9.

C: > svrmgr30 C: > connect internal/password 10. Start the database in the NOMOUNT state: SVRMGR> STARTUP NOMOUNT PFILE=c:\mars\initmars.ora 11. Turn on spooling to trap error messages and run BUILD_MARS.SQL: SVRMGR> SPOOL build.log SVRMGR> @BUILD_MARS.SQL If there are errors while running BUILD_MARS.SQL, fix the errors and rerun the script for successful completion. 12. Generate the data dictionary by running CATALOG.SQL: SVRMGR> @%RDBMS80%\ADMIN\CATALOG.SQL 13. Run CATPROC.SQL to generate the objects used by PL/SQL: SVRMGR> @%RDBMS80%\ADMIN\CATPROC.SQL 14. If you want additional features, run the appropriate scripts, such as CATREP8M.SQL for Advanced Replication. 15. Turn off spooling and check the log for errors. All the MAX parameters are set when the database is created. To determine what parameters your database has been created with, execute the following: SVRMGR> Alter database backup controlfile to trace This command will create an SQL script that contains several database commands: CREATE CONTROLFILE REUSE DATABASE "SJR" NORESETLOGS NOARCHIVELOG MAXLOGFILES 32 MAXLOGMEMBERS 2 MAXDATAFILES 254 MAXINSTANCES 1

http://www.informit.com/content/0789716534/element_002.shtml (9 of 13) [26.05.2000 16:46:14]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

MAXLOGHISTORY 899 LOGFILE GROUP 1 'E:\ORANT\DATABASE\LOGSJR1.ORA' GROUP 2 'E:\ORANT\DATABASE\LOGSJR2.ORA' DATAFILE 'E:\ORANT\DATABASE\SYS1SJR.ORA', 'E:\ORANT\DATABASE\RBS1SJR.ORA', 'E:\ORANT\DATABASE\USR1SJR.ORA', 'E:\ORANT\DATABASE\TMP1SJR.ORA', 'E:\ORANT\DATABASE\INDX1SJR.ORA' ;

SIZE 200K, SIZE 200K

To generate SQL statements for all the objects in the database, Export must query the data dictionary to find the relevant information about each object. Export uses the view definitions in CATEXP.SQL to get the information it needs. Run this script while connected as SYS or "internal." The views created by CATEXP.SQL are also used by the Import utility. Chapter 25, "Using SQL*Loader and Export/Import," discusses more about Oracle's Export and Import utilities. CATALOG.SQL and CATEXP.SQL views don't depend on each other You don't need to run CATALOG.SQL before running CATEXP.SQL, even though CATEXP.SQL is called from within CATALOG.SQL. This is because no view in CATEXP.SQL depends on views defined in CATALOG.SQL. Create an identical copy of database but with no data 1. Do a full database export with ROWS=N: C: > exp system/manager full=y rows=n file=fullexp.dmp This will create a full database export (full=y) without any rows (rows=n). 2. Run a full database import with ROWS=N: C: > imp system/manager full=y rows=n file=fullexp.dmp Creating a new database on the same machine If the new database is to be created on the same machine as the old database, you need to pre-create the new tablespaces because the old data files are already in use. Use Instance Manager to create a new database in Windows NT 1. From the Start menu choose Oracle for Windows NT and then NT Instance Manager. This will start the Instance Manager and show you the status and startup mode of all the SIDs (see Figure 2.1). Figure 2.1 : The Instance Manager dialog box shows the available instances. 2. Click the New button and supply the SID, internal password, and startup specifications for the new instance (see Figure 2.2). Figure 2.2 : Provide the specifications for the new instance. 3. Click the Advanced button and choose appropriate database name, logfile, and data file parameters and a character set for the new database (see Figure 2.3). Figure 2.3 : Provide the specifications for the new database. The Oracle Database Assistant can be used to create a database at any time. Use Oracle Database Assistant to create a new database in Windows NT 1. From the Start menu choose Programs, Oracle for Windows NT, Oracle Database Assistant. 2. Select Create a Database and click Next. 3. Choose the Typical or Custom option and click Next. The Custom option lets you to customize the parameters of the database that you're trying to create. 4. Choose Finish. In Windows NT, you can set the default SID by setting the Registry entry ORACLE_SID. Updating ORACLE_SID in the Windows NT Registry 1. From the DOS command prompt, type REGEDT32. Don't modify the Registry unless you know what you're doing!

http://www.informit.com/content/0789716534/element_002.shtml (10 of 13) [26.05.2000 16:46:14]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

Be extremely careful when working with the Registry. Improperly set keys may prevent Windows NT from booting up. 2. Choose the key \HKEY_LOCAL_MACHINE\ SOFTWARE\ORACLE\HOMEID. 3. From the Edit menu choose Add Value. 4. In the Value Name text box, type ORACLE_SID. 5. For the Data Type, choose REG_EXPAND_SZ. 6. Click OK. 7. Type your SID name in the String Editor text box and click OK. 8. Exit the Registry.

Checking the Status of your Database

After the database is created, regularly check the status of the database by examining its data dictionary and the alert log.

Examining Basic Views

The data dictionary is one of the most important parts of the Oracle database. The data dictionary is a set of tables and views that you can use to look up valuable information about the database. You can use the data dictionary to obtain various types of information, including: q General database structure q Information about schema objects q Integrity constraints q Database users q Privileges and roles q Space allocated to database objects The catalog.sql and catproc.sql scripts can be used during or after database creation to create the commonly used data dictionary views and for PL/SQL support, respectively. The data dictionary contains a set of base tables and associated set of views that can be placed in the following categories: View Category Description Views accessible by any user that provide information on objects owned by them USER_xxx Views accessible by any user that provide information on all objects accessible by ALL_xxx them Views accessible by any user that provide information on any database object DBA_xxx Which data dictionary objects do I have? All the data dictionary tables and views are owned by SYS. You can query the DICTIONARY table to obtain the list of all dictionary views. The following examples show how to query the dictionary tables to obtain information about the database: q To identify all the rollback segments in the current database and their status, use Select * from dba_rollback_segs; To identify all the data files in the current database and their status, use Select * from dba_data_files; To identify all the tablespaces in the current database and their status, use Select * from dba_tablespaces; To identify all the users belonging to this database, use Select * from dba_users;
http://www.informit.com/content/0789716534/element_002.shtml (11 of 13) [26.05.2000 16:46:15]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8
q

To find out if the database is in the ARCHIVELOG mode, use Select * from v$database; To identify all the parameter values in use for the data-base, use Select * from v$parameter;

Checking the Oracle Alert Log

When diagnosing a database problem, the first place to look for information and errors is the alert log (the name is operating system dependent). If this file isn't present, Oracle will automatically create it during database startup. This file can point you to the location of trace files, which can give a lot of insight into the problems encountered. It also contains additional information to indicate the status of the database and what's now happening in the database. Locating trace files The trace file would be located in the directory specified by BACKGROUND_DUMP_DEST, USER_DUMP_DEST, or CORE_DUMP_DEST, depending on the exact error and its cause. SEE ALSO For more information on the contents and usage of the alert log, When the database is started, the following information is recorded in the alert log: q All the init.ora parameters q Informational messages indicating that the background processes have been started q The thread used by the instance q The log sequence that the LGWR is now writing to In general, the alert log records all important incidents of the database, including: q Database startups q Database shutdowns q Rollback segment creations q Tablespace creations q ALTER statements issued q Log switches q Error messages Each entry has a timestamp associated with it, and each non-error message has an entry marking its beginning and another entry marking its successful completion. You should frequently check this file for error messages for which the alert log will point to a trace file for more information. The following is a sample alert log:

File header showing information about your system Initialization parameters Database in nomount state and CREATE DATABASE command.

LOGFILE 'E:\ORANT\database\logSJR1.ora' SIZE 200K, 'E:\ORANT\database\logSJR2.ora' SIZE 200K MAXLOGFILES 32 MAXLOGMEMBERS 2 MAXLOGHISTORY 1 DATAFILE 'E:\ORANT\database\Sys1SJR.ora' SIZE 50M MAXDATAFILES 254 MAXINSTANCES 1 CHARACTER SET WE8ISO8859P1
http://www.informit.com/content/0789716534/element_002.shtml (12 of 13) [26.05.2000 16:46:15]

informit.com -- Your Brain is Hungry. InformIT - Creating a Database From: Using Oracle8

NATIONAL CHARACTER SET WE8ISO8859P1 Thu Jan 29 09:33:50 1998 Successful mount of redo thread 1. Thread 1 opened at log sequence 1 Current log# 1 seq# 1 mem# 0: E:\ORANT\DATABASE\LOGSJR1.ORA Successful open of redo thread 1. Thu Jan 29 09:33:50 1998 SMON: enabling cache recovery Thu Jan 29 09:33:50 1998 create tablespace SYSTEM datafile 'E:\ORANT\database\Sys1SJR.ora' SIZE 50M default storage (initial 10K next 10K) online Thu Jan 29 09:34:10 1998 Completed: create tablespace SYSTEM datafile 'E:\ORANT\datab Thu Jan 29 09:34:10 1998 create rollback segment SYSTEM tablespace SYSTEM storage (initial 50K next 50K) Completed: create rollback segment SYSTEM tablespace SYSTEM Thu Jan 29 09:34:14 1998 Thread 1 advanced to log sequence 2 Current log# 2 seq# 2 mem# 0: E:\ORANT\DATABASE\LOGSJR2.ORA Thread 1 cannot allocate new log, sequence 3 Checkpoint not complete Current log# 2 seq# 2 mem# 0: E:\ORANT\DATABASE\LOGSJR2.ORA Thread 1 advanced to log sequence 3 Current log# 1 seq# 3 mem# 0: E:\ORANT\DATABASE\LOGSJR1.ORA Thread 1 advanced to log sequence 4 Current log# 2 seq# 4 mem# 0: E:\ORANT\DATABASE\LOGSJR2.ORA < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_002.shtml (13 of 13) [26.05.2000 16:46:15]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Migrating an Oracle7 Database to Oracle8

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q

Why Migrate? Selecting a Migration Method

r r r

Selecting Oracle's Migration Utility Using Export/Import Using Table Copying Identifying Types of Tests Setting Up a Test Program Testing and Retesting Executing the Migration Process with Oracle's Migration Utility Executing the Migration with Export/Import or Table Copying Precompiler Applications OCI Applications SQL*Plus Scripts SQL*Net Enterprise Backup Utility (EBU) Standby Databases

Testing the Migration

r r r

Performing the Migration

r r

Completing Post-Migration Steps

r r r r r r

Migration: Final Considerations

Why Migrate?
You may want to migrate an Oracle7 database to Oracle8 for a number of reasons. You may want to take advantage of one or more of Oracle8's new features, outlined in Appendix B, "What's New to Oracle8." You may simply want to benefit from the faster processing that the revised code tree should allow. Whatever the reason, you have a number of options regarding the method you can use to complete the migration process. One of them is a migration tool provided by Oracle. Although this chapter concentrates on the migration tool, it also discusses the alternatives. In the following section you learn about all the options. After reading it, you should be able to determine which method is best to use to migrate your database. The structural changes in Oracle8

http://www.informit.com/content/0789716534/element_003.shtml (1 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

A migration is necessary because the new functionality in Oracle8 requires changes to the basic items in the data dictionary. Until the new dictionary is built, the Oracle8 kernel can't operate successfully. In addition, the structure of the data file header blocks has changed to support some of the new features. These changes must be in place for the code to work correctly. Unlike simple upgrades such as those you might have performed to move from version 7.2 to version 7.3, these structural changes require more than simply installing the new code and relinking the applications. For more details and a further discussion of the migration options you should read the Oracle8 Server Migration manual, part number A54650-01.

Selecting a Migration Method

The end result of a migration from Oracle7 is a database that contains essentially the same user objects as the original database, but in data files with updated headers and supported by a data dictionary that allows the new Oracle8 features. In some cases, this may not be your ultimate goal. For example, you might want to migrate only a portion of your database for testing purposes, where you want to test only a subset of your applications before migrating the entire database, or because you don't need objects created for now-obsolete portions of the application. On the other hand, you might want to use the downtime required for the migration to make some structural changes to the database. This might include moving segments between different tablespaces or may simply involve coalescing free space in one or more fragmented tablespaces. We will examine three basic approaches to migration in this chapter: Oracle's Migration utility, export/import, and table copying. You can read the details concerning each strategy in the following sections and decide which best suits you. To get started, look at Table 3.1 to see the basic features of each approach. Choose your best migration method Your choice of migration method will depend on how much you want to accomplish as part of the migration and on how much space you have to complete the task. It might also depend on the length of time you can afford to make the database inaccessible, because some methods take much longer than others. TABLE 3.1 Overview of migration options Option: Migration utility Export/Import Table copying Migrate Only: Yes No No Need for Additional Space: System tablespace Export dump file Two databases Time Requirements: Least Great Greatest

As you can see, the fastest approach, the one needing the least overhead, is the Migration utility. However, you can't include other database-restructuring or related changes if you use this method. The Migration utility will migrate your entire database as it is. With the two other options, you can make changes to the structure, layout, and tablespace assignments, but you'll need more time and disk resources to complete these tasks. They're also more complicated to complete because you need to perform a number of additional steps. The details of the steps needed to complete each type of migration (and the reasons for choosing each) are listed in the appropriate sections following. Table 3.2 summarizes these options. TABLE 3.2 Summary of migration method characteristics Migration Utility: Automatic: Requires little DBA intervention Requires minimal extra disk space Time is factor of number of objects, not database size Can only migrate forward Can't use for release to release All or nothing Export/Import: Requires a new database build Can use large amounts of disk space Very slow for large databases Can migrate forward and backward Can use for release to release Partial migration possible Copy Commands: Requires lots of attention Requires both databases to be online Very slow for large databases Can migrate forward and backward Can use for release to release Partial migration possible

http://www.informit.com/content/0789716534/element_003.shtml (2 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

No structural changes can be made

Concurrent defragmentation and reorganization

Concurrent defragmentation and reorganization

Selecting Oracle's Migration Utility

You need to consider several key factors when planning to use the Migration utility: q Available space in the system tablespace q Need to migrate only a subset of the database q Resources to test a full database migration q Ancillary requirements, such as space defragmentation Oracle's Migration utility is designed to perform the required structural changes to your existing database. It will actually build a new Oracle8 data dictionary in the same system tablespace as your current Oracle7 dictionary, and it will restructure your rollback segments and the header blocks of the database's data files on disk (see Figure 3.1). Users' objects, such as tables and indexes, stay just as they are, although the ways in which they're accessed-through the data dictionary and then to the files where they really reside-are changed. Again, the new internal structure is designed to make the database more efficient and to support a new set of features. Figure 3.1 : The Migration utility changes database structures, including the data dictionary, rollback segments, and data file header blocks, in place. The migration process requires that your current Oracle7 and your new Oracle8 data dictionary reside in the database for a short period of time. This means that the system tablespace, the dictionary's home, must be large enough to hold both versions simultaneously. Therefore, the first item you need to consider before deciding whether to use the utility is the space you have available for the system tablespace. Your Oracle8 data dictionary will be about 50 percent larger than your Oracle7 dictionary. Of course, you may already have some of this space available, but most DBAs will find they need to add more. Space requirements for the Migration utility For a period of time, you'll need to have space for two versions of the data dictionary in the system tablespace and for two releases of Oracle in their respective Oracle Home directory structures. The data dictionary will require about 2 1/2 times the space now consumed in your system tablespace. The Oracle8 installation will take 500MB-more if you select many options. If you don't have the required disk space, either consider another migration strategy or wait until you can add the required capacity. Your next decision point for using, or not using, the Migration utility is whether you want to migrate the database as is, or if you also want to make some changes. The Migration utility is an all-or-nothing tool. You can't migrate portions of the database because the database is migrated in situ. Similarly, because the data isn't being moved, you can't move segments between tablespaces and you can't coalesce free space in your tablespaces as part of the migration process. Some DBAs don't like to make too many changes at one time, so even if they want to complete some restructuring tasks, they'll make these changes independently of the migration, completing them before or after the migration itself. Others have limited time windows in which to complete maintenance work and so try to make all required changes at a single time. You need to consider what else, if anything, you want to achieve as part of the migration processing. You also need to consider how much time you can take away from your user community while working on these steps. Consider test options for a migrated database If you want to convert your data-base to perform functional and similar tests, you can't use the Migration utility for continued work under Oracle7. If you make a copy of it first, you can convert the copy and simultaneously run the production Oracle7 database and the test Oracle8 version. However, making a copy is time-consuming, usually accomplished with the Export/ Import tools or with some form of data unloader and SQL*Loader. You can create a partial test database by using the tables for just one or two representative application tasks. You'd still need to complete integrated tests (if you needed them) after a full database conversion. If you only need to perform a migration, Oracle's utility is a good choice. First, it's relatively fast due to the changes being made on the current database structures. They don't have to be copied, moved, or otherwise duplicated-all relatively slow processes. In addition, the only factors that really affect the migration speed are the size of the System tablespace and the number of data files. The System tablespace is typically a small

http://www.informit.com/content/0789716534/element_003.shtml (3 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

portion of the overall database, and the number of data files is limited to 1,022 in Oracle7. Thus, even the largest databases typically take no more than a day to migrate. Piecewise migration with Export/Import Using this technique to migrate your database one piece at a time requires you to keep both database versions available, which means maintaining the Oracle7 and Oracle8 executables online. Further complications from this approach occur if the data in the different versions is in any way related. You might need to have users switch between databases to perform different functions or temporarily build a distributed database environment. It may also require that parts of both databases be inactive when it's time to move additional segments from Oracle7 to Oracle8. If you're moving only part of your database because you don't need the rest of it, these issues become irrelevant. Later in this chapter's "Executing the Migration Process with Oracle's Migration Utility" section you'll find a detailed description of how to complete a migration with the Migration utility. First look at the other migration options and the test plan you need to construct, regardless of the migration approach you'll take.

Using Export/Import
If you decide to use the Export/Import tools to migrate your database, you need to plan for the following resources to be available: q Space to store the export file q Time to create the export q A copy of the Oracle8 executables q An empty Oracle8 database for the file being exported q Time to perform the import The amount of space and time needed for the initial export depends on the amount of data being exported. If you decide to move only part of your database to Oracle8, you need less time than if you are transferring the entire database. The time also depends on the speed of the devices to which you export. A fast disk drive allows a faster export than a slower tape drive. A very large database may also require a file too large for the operating system or for Oracle to handle. In this case you may need to use some form of operating system tool, such as a pipe, to move the data onto the appropriate media. Figure 3.2 shows the typical export/import steps. By using a pipe, you can send the output from your Export directly to the Import utility's input. Figure 3.2 : Migrating your database with export/import requires two distinct steps; the export dump file is the intermediary. Migrate your database via export/import 1. Perform a full database export from Oracle7, after which you can remove the Oracle7 database and the Oracle7 home directory structure. 2. Install Oracle8 and then alter the environment variables and your parameter file to point to the Oracle8 structures. (See Appendix C, "Installing Oracle8," for installation instructions.) 3. Create an Oracle8 database. 4. Add the required tablespaces to this database. 5. Perform a full import. Protect your current database Before beginning your migration, it's recommended that you ensure you have a backup of the home directory and of the database. Minimally, it's recommended that you keep the scripts you used to build the database in the first place; that way you have at least one easy way to reconstruct the database in case you run into problems with the Oracle8 version. SEE ALSO To learn how to create an Oracle8 database, To add the required tablespaces to the database, A variant of this method is to use an unload/loader approach to move the data. You can do this by building your own unloader utility or by finding one in the public domain. An unloader utility needs to extract the rows of data from your tables, as well as the definitions of the tables and all the other database objects; that includes indexes, userids, stored procedures, synonyms, and so on. You can also consider a hybrid approach, using the export to create only the object definitions and the unloader simply to create the row entries.

http://www.informit.com/content/0789716534/element_003.shtml (4 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

Advantages of unloader/loader technique The big advantage to the unloader/loader approach is that you can use Oracle's SQL*Loader utility to reinsert the data when the definitions are applied to the data-base. This utility, running in its direct path mode or-even better if you have the hardware to support it-in parallel direct mode, can complete the job of loading records much more quickly than the Import program. SEE ALSO To learn more about the Export and Import utilities, To learn more about SQL*Loader,

Using Table Copying

As with export/import, you can use table copying to move just part of your database, to stage the migration, or simply to avoid migrating unneeded elements. The same caveats for incomplete database migration apply in this context as for the export/import approach. To perform table copying, you need to have both databases (Oracle7 and Oracle8) available simultaneously, which includes not just the database storage but the two Oracle Home structures and the environments to run them both simultaneously. This makes the approach the most space-intensive of all three methods (see Figure 3.3). Figure 3.3 : Both databases must remain online during a migration using table copying. The other drawback to this approach is that it really does only copy table definitions and their contents. To build the same database in the Oracle8 environment that you started with-including all the users, stored procedures, synonyms, views, and so on-you still need to find a method to copy these from Oracle7. It's therefore likely that you'll need to perform a partial export/import or even a data unload/reload, as discussed in the preceding section. General steps for migrating from Oracle7 to Oracle8 1. Install Oracle8. 2. Create an Oracle8 database. 3. Add the required tablespaces and users. 4. Select the Oracle7 or Oracle8 database as your primary database. 5. Configure SQL*Net or Net8 with a listener to connect to your secondary database. 6. Create or modify a TNSNAMES.ORA file to identify your secondary database. 7. Use SQL*Plus to issue the required COPY commands for your primary database. 8. Add the views, synonyms, stored procedures, and other objects dependent on the tables into your secondary database. 9. Drop your Oracle7 environment. Warning: Avoid points of no return As always, before moving or destroying your production system, you should make a backup first. Copying across database links Rather than use the SQL*Plus COPY command, you can complete step 7 by creating data-base links in your primary database to access the secondary database and by using SQL CREATE TABLE...AS SELECT commands to copy table definitions and data between your database. You find a detailed discussion of these steps later in the section "Executing the Migration with Export/Import or Table Copying."

Testing the Migration

Software development and maintenance efforts should always include a good test plan as part of the acceptance strategy. Your migration from Oracle7 to Oracle8 is no different. Indeed, because you already have a working production database, you need to devise a test strategy to ensure that the end results of the migration contain the same, or better, production capabilities. This means testing not only the capabilities, but the performance and results as well.
http://www.informit.com/content/0789716534/element_003.shtml (5 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

You can apply a number of types of tests to help assure you and the users that everything is working properly at the end of the migration. In the next sections you see what types of tests you can use and when to use each type. From these selections, you can build a test program and identify the resources needed to complete the tests. Oracle Corporation strongly recommends running all these tests before concluding the migration.

Identifying Types of Tests

You can perform six basic types of tests to validate the migration: q Migration test q Minimal test q Functional test q Integration test q Performance test q Volume/load stress test Migration Test This test validates your migration strategy, whether it's to use the Migration utility or one approach to transfer the data from Oracle7 to Oracle8. It's intended to help you determine whether you've allotted sufficient resources for the migration, including disk space, time, and personnel. Running a migration test 1. Create either a test copy of your whole database, or a subset if you don't have the resources to test the complete database migration. You don't have to complete this step if you're using export/import or table copying. 2. Execute the migration with the utility you plan to use when migrating your production database. 3. Upgrade the tools used by your application. Resolving problems with migration tests If this test fails, you may need to rethink the chosen migration strategy and possibly choose another method. For example, if the time taken is greater than your users are willing to allow, you may need to plan a staged migration, or if you were hoping to use export/import, you may need to consider the Migration utility instead. Minimal Test This type of testing involves migrating all or part of an application and simply attempting to run it. No changes are made to the application, and performance and value testing should be attempted. This test simply confirms that the application can be started against the migrated database; it's not intended to reveal all problems that could occur. Resolving problems with minimal tests A failure of a minimal test typically indicates problems with the migration itself, such as missing tables, synonyms, views, or stored procedures. It's most likely to fail if you've tried a partial database migration or if you're using one of the data-transfer strategies rather than the Migration utility. You should perform this test after a successful migration test and before moving on to the more rigorous tests. It will require relinking the tools used by the applications, so you'll need to maintain a separate copy of the application if the users need to continue using the production Oracle7 database. Running a minimal test 1. Complete the migration test. 2. Have users, developers, or a test suite run selected programs from the applications. Functional Test The functional test follows the minimal test and ensures that the application runs just as it did before the migration. This involves having users, or simulated users, executing the different application components and verifying that the outcome is the same as in the pre-migrated database. The results of any queries, reports, or DML should be the same as they were on the pre-migrated database.

http://www.informit.com/content/0789716534/element_003.shtml (6 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

If you're using Oracle8 to enhance the application, you may also want to add the new functionality to each application during this test phase to ensure that the application continues to provide reliable results with the new features in place. Conducting a functional test 1. Complete the migration and minimal tests. 2. If you intend to add new functionality to your applications, have the developers make these changes. 3. Have users, developers, or a test suite execute the applications, testing all functions and features. Tracking the cause of errors detected during functional testing may involve close cooperation between the DBA and the application developers. It's important, therefore, to ensure that the development organization is apprised of this testing phase and can commit the necessary resources. If you're running third-party application software, you may need to get help from your vendor should this test fail. Testing third-party applications Some vendors may not be aware of all the changes made in Oracle8. If you're using third-party applications, you shouldn't commit to a completed migration until the functional tests have been rigorously completed. Integration Test Integrated testing involves executing the application just as you did in the pre-migrated database. This includes establishing client/server connections, using any GUI interfaces, and executing testing online and batch functions. This test ensures that all the application's components continue to work together as before. Resolving problems with integration tests Should you run into problems with these tests, you'll have to isolate whether the cause is in a single component, such as SQL*Net or Net8, or whether it's part of the overall migration. Generally, if you've completed the functional testing successfully, the likelihood is that the problem is with one component, or the interface between a pair of components. Running an integration test 1. Complete the migration, minimal, and functional tests. 2. Install and configure any communication software, such as Net8, for client/server or multi-tier architectures. 3. Install and configure any drivers, such as ODBC drivers, that the applications use. 4. Have the users, developers, or a test suite run the applications across the network, using the same front-end tools and middleware that are now planned for database access. Performance Test Although the kernel code tree has been optimized in Oracle8, you might discover that some parts of your applications aren't running as well as before the migration. This could be due to a number of factors, such as tuning efforts that were made to avoid a problem in the earlier release. You need to run the performance tests to ensure that overall processing throughput is at least the same as, if not better than, the Oracle7 performance. Resolving problems with performance tests If you find performance problems, you should attempt to resolve them by using the database tuning techniques described in Chapter 20, "Tuning Your Memory Structures and File Access," through Chapter 23, "Diagnosing and Correcting Problems." Conducting a performance test 1. Complete the previous tests to ensure that you're running the equivalent of a full production system. 2. Have users run their interactive and batch programs as they would in a production environment. 3. Monitor and record the database performance by using queries against the various dynamic performance tables or by using such tools as the UTLBSTAT.SQL and UTLESTAT.SQL scripts. 4. Solicit feedback from users as to their perceptions of performance and response times compared to the current production system. If you've been monitoring your Oracle7 database with the various analytic and diagnostic tools, you can easily make comparisons by using the same tools on the migrated database. SEE ALSO
http://www.informit.com/content/0789716534/element_003.shtml (7 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

For an overview of the dynamic performance tables, A detailed description of the UTLBSTAT.SQL and ULTESTAT.SQL utilities begins on Volume/Load Stress Test Ideally, you should be able to test your migrated database against a realistic workload. This includes the amount of data being processed (volume) and the concurrent demands on database (load). To perform such testing, you may need to set up automated procedures rather than expect your user community to test your database under realistic conditions while continuing work on the unmigrated production version. This test will ensure that the database is ready for the workload intended for it and should also display any other problems that the other tests didn't uncover. Performing volume/load stress tests 1. Assemble either a workforce or automated scripts to represent a normal, everyday workload. 2. Exercise the system by having the users or scripts execute the applications concurrently. 3. Monitor and record the system performance as in the performance testing. Building a load test If you have software that can capture the keystrokes entered during an interactive session, you can use this to collect the session work completed by the users in earlier tests. You can use these to build scripts that emulate those sessions. Run multiple concurrent copies of these scripts to simulate different levels of system load. Due to changes in the structure and use of internal structures-the data dictionary, rollback segments, and ROWIDs-you may find that the behavior of the database changes differently from the way it did in Oracle7. Although most resources won't reach a performance threshold as quickly as they might in Oracle7, you can't depend on this. It's therefore not advisable to assume that if you achieve performance equal to or better than Oracle7 with a small number of concurrent sessions manipulating a few tables, this performance level will be maintained under full volume and load. Addressing problems with a volume/load stress test Problems encountered while testing for volume and load should be addressed by applying the tuning strategies discussed in Chapters 20 through 23 of this book.

Setting Up a Test Program

Keeping in mind the various tests you need to perform, your test program should address the when, where, what, who, and how questions associated with each test. The test program should also address the methods you'll use to compare the actual results with what should be expected if the test is successful. This may include creating test suites that can be run on the current production database and on the Oracle8 test database; on the other hand, it could include simply recording the sizes of such objects as temporary segments and rollback segments in the test database so that you'll be prepared to size the associated tablespaces appropriately when the migration is performed for real. When to perform the tests depends on the resources you need and their availability. For example, DBAs are used to working on major database changes during periods of low activity, such as late at night, weekends, and holidays. If your test needs the participation of developers or end users, however, you may have to plan the test during normal working hours. Ensure the integrity of your test environment There's no point performing a test for validity after migration if the original version is flawed. Similarly, if you plan to test just part of your database, you need to ensure that you'll get a valid subset of the data. For example, if you're going to test a function that adds new records to a table and a sequence generator is used for the primary key values, you'll have to ensure that the sequence generator is avail-able in the pre-migration set of objects. Where to perform your tests depends on your computer environment. Ideally, you want to test as much of the database as you can-all of it if possible. This may require using a separate machine if one is available. A test or development machine is probably the best place to run the various tests. Remember, however, that you may have to schedule this machine if it's regularly used by the developers; some tests may require shutting down the Oracle7 database(s) running there. What to test depends on the test you're performing. By referring to the previous section's descriptions of the different tests, make sure that you have the resources to complete the test and record the findings in a

http://www.informit.com/content/0789716534/element_003.shtml (8 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

meaningful way. For example, you won't learn anything about whether the Oracle8 performance is equal to, better than, or even worse than the Oracle7 performance if you don't have a method to record the performance characteristics you want to measure in each environment. Who to involve in the testing also depends on the type of test. The earlier test descriptions should help you identify the type of personnel needed for each one. You may want to form a migration team with members from your system support, developer, and end-user communities if you're going to migrate a large database. This team can help you schedule the tests in such a way that they don't cause major conflicts with other groups. For example, you would want to avoid running a test that needs input from the users during times of heavy workloads, such as month-end processing. The team can also help you find the best resources within their respective groups to aid with the tests and can act as your communication channel back to the various groups regarding the migration progress. How to complete the tests depends on your environment as well as the test type. You need to decide if you'll run tests on the whole database or on partial applications. This, of course, depends on the resources you have available. Similarly, you need to ensure that you have the resources, including people and tools, to fix any problems encountered during the testing so that you can keep the migration project on track. The individuals needed to fix a problem may not be the same as those involved in the test itself. The how question needs to include how you'll obtain your test data. If you want to test against the entire database, you'll need a method to create an exact copy of it, possibly on a separa ate machine. This could involve an export/import or some form of Unload/Reload utility. If using the latter, you need a verification test suite to ensure that the copy was successful. After your test plan is in place, you can begin the process of fully testing a migration. Ideally, you'll run every test on a complete test version of the migrated database before tackling the migration of the production system.

Testing and Retesting

As you complete each test in your test plan, you should be able to determine whether it's successful. If the test is successful, you can move to the next one; if it isn't, you need to fix the problem and retry the test. This part of the testing isn't always as straightforward as it sounds. Suppose you encounter an error that involves a missing view that should contain the join of two tables. The error could reflect that one of the two underlying tables is missing, or that the view definition is no longer available or valid. If you've performed a full migration, you need to determine whether the Migration utility "lost" the view or table or whether the view (or table) was missing before the migration was performed. If you don't have a copy of the pre-migrated database (or at least a way to reconstruct it), you can't determine the cause of the error. You'll have to go back and redo all the steps you took to get to this testing point, which may include copying the current production database over to your testing environment. Of course, because it has been in active use since you made your initial copy, the current copy you make won't be the same as the one you used for the test that failed. For example, the view may now have been deliberately dropped. You'll have to repeat all the tests to validate this new version of the database. If you're testing a subset of the database, the missing view or table may not have been created because it wasn't included in the objects selected for migration. In this case, you need to decide whether you can just add it now and continue with your testing. Otherwise, as in the preceding case, you'll need to start over with the migration of the test set from an Oracle7 source and repeat all the testing. If you run into a problem that you can't easily resolve, you have two options: q You can continue with further tests to see if a more obvious reason for the problem manifests itself. If you still can't determine the cause of the problem at the conclusion of your tests, you should repeat the entire migration process up to the point where the problem was observed. This time, make detailed notes or keep log files of all your activities so that you can file a detailed report with the Oracle Support organization, should the problem recur. q Simply restart the process as just discussed, without continuing further testing; keep detailed records in case you need to file a request for help with Oracle Support.

Performing the Migration

After you complete an acceptable test plan, you should use it to migrate and test a non-production version of your database. When you're certain that you're ready, you can perform your production system's migration. Skip sections that don't relate to your chosen migration approach

http://www.informit.com/content/0789716534/element_003.shtml (9 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

If you're planning to use the Migration utility, continue with the following section. If you intend to use export/ import for your migration, skip to "Executing the Migration with Export/Import or Table Copying" in this chapter.

Executing the Migration Process with Oracle's Migration Utility

The database is addressed as a whole in the following, detailed descriptions of the tasks you'll have to perform to complete your database migration. If you need to migrate only a portion of your production database, you must create a temporary Oracle7 database to hold just that portion. Apply the migration processing to this temporary database only. General steps for preparing to migrate with the Migration utility 1. Load the Migration utility by using the Oracle8 installer. 2. Ensure that you have sufficient free space in the system tablespace. 3. Recover or remove any offline objects. 4. Remove any user with an ID of MIGRATE. 5. Override any in-doubt transactions. After you complete the first task (loading the Migration utility), you may want to put your database into restricted mode to prevent users from making unwanted changes as you prepare it for the migration. To do this, perform a shutdown and then reopen it with the RESTRICTED option. This will disconnect all current users and allow only those with the restricted session privilege to reconnect. If you're one of many DBAs with such a privilege, you should coordinate with your colleagues to ensure that only one of you is working on the migration process. SEE ALSO For details on the various options for starting a database, including the RESTRICTED option, Loading the Migration Utility The Migration utility for converting from Oracle7 to Oracle8 is provided as part of the Oracle8 installation media. You can load just the Migration utility by running the standard installation program for your specific hardware platform (the orainst program in UNIX and the SETUP.EXE program in Windows NT, for example). Suggested responses to the installer's questions 1. Choose Install, Upgrade, or De-install on the Select the Installer Activity screen. 2. Choose Migrate from Oracle7 to Oracle8 on the Select Installer Option screen. 3. Choose Install Migration Utility on the Select an Oracle7 to Oracle8 Migration Option screen. The installer will place a number of files into the Oracle7 home directory structure, including the following: q The Oracle8 Migration utility, placed in the bin sub-directory q The Oracle8 version of the message file, placed in the MESG subdirectory of the RDBMS subdirectory q The Oracle8 version of the MIGRATE.BSQ file, placed in the DBS subdirectory q Any required NLS files, placed in a subdirectory named data, which is in the path from the Oracle home directory that contains the subdirectory tree MIGRATE, NLS, and ADMIN. Confirm success of your installation Following the installation, you should check the log file to confirm that the files were successfully installed. Checking for Sufficient Space in the System Tablespace The Migration utility needs space for the Oracle7 and Oracle8 data dictionaries in the system tablespace. You can determine whether you have sufficient space available by using a special option in the Migrate utility. Simply run the utility with the CHECK_ONLY option set to TRUE: q On UNIX, the command is mig check_only=true On Windows NT, the command is

http://www.informit.com/content/0789716534/element_003.shtml (10 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

mig80 check_only=true Depending on your operating system, the name of the utility and the format of the results will vary. You'll typically need free space equivalent to about 1 1/2 times the space consumed by your current data dictionary. Confirming That No Tablespaces or Data Files Need Recovery All offline tablespaces should be brought back online unless you're certain that they were taken offline by using the TEMPORARY or IMMEDIATE option. After you bring them back online, you can use one of these options to take them back offline. Unusable tablespaces If you can't bring a tablespace back online because it needs recovery that can't be completed, you need to drop it; it will be unusable under Oracle8 anyway. All data files must also be online. You can check the DBA_DATA_FILES view for the status. If any are offline and you can't bring them back online because they need recovery, the Migration utility will fail with errors. Don't have a user called MIGRATE The migration process will create a user called MIGRATE. Because this user is eventually dropped with the Oracle7 data dictionary objects, you should ensure that you don't already have a database user with this name. If you do, create a new schema to contain the MIGRATE user's objects, or use a user-level export and plan to reimport the user following the migration. In either case, remember to drop the MIGRATE user after you save the objects from the schema. See Chapter 9 "Creating and Managing User Accounts," for information about user and schema management. SEE ALSO For a brief discussion of views, Ensuring That You Don't Have Any Pending In-Doubt Transactions If you've used distributed transactions in your Oracle7 database, you need to check that none are still pending due to problems with the two-phase commit mechanism, such as lost network connections or offline databases. You can find such transactions by examining the DBA_2PC_PENDING table. If you have any such transactions, you need to commit or roll them back manually. You can find the instructions on how to do this in your Distributed Database documentation, including details on how to determine if you should commit or roll back. Performing a Normal Shutdown of the Oracle7 Database When you've readied your database for the migration by performing the preceding tasks, you can shut down your database. You need to shut it down cleanly-that is, with the NORMAL or IMMEDIATE option. If you can't do this and have to use the ABORT option, you need to restart the database and then shut it down again with one of the other options. This ensures that there are no pending transactions or incomplete checkpoints, leaving your database in the appropriate state for the migration. SEE ALSO For details on database shutdown options and commands, Backing Up the Database in Case of Problems After your database is shut down, you should make a full backup just in case the migration process needs to be repeated, as discussed in the earlier section on testing. The backup needs to be made any time you plan to migrate a database that has been opened subsequent to your last pre-migration backup-unless you don't mind losing the changes made during that period. Hot backup option before migration If you don't have the time to complete an offline backup, you can complete an online backup immediately before shutting it down for the migration. Remember that as soon as it's closed, you should back up the online redo logs as well. If you need to restore the Oracle7 version for another migration attempt, you have to recover the backup to a stable point, which requires the contents of the online redo. SEE ALSO For an overview of hot backup strategies,

http://www.informit.com/content/0789716534/element_003.shtml (11 of 16) [26.05.2000 16:46:26]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

Detailed descriptions of hot backup steps are available on Run the Migration Utility You may need to set certain system values before running the Migration utility program. These will vary between operating systems, and you need to examine your platform-specific documentation for details on what to set and what values they require. For example, the TWO_TASK and ORA_NLS33 variables have to be set appropriately. You also need to use this documentation to find out how to run the migration program and provide the appropriate options. The options for the migration program are documented in Table 3.3. Table 3.3 Options for the Migration Program Name: CHECK_ONLY or NO_SPACE_CHECK Description and Use: These mutually exclusive options are used to determine whether the System tablespace is large enough to complete the migration or to avoid making this check. You should need the CHECK_ONLY option only in the premigration steps, as discussed earlier. This option specifies the name of the database to migrate. This option specifies the new name for the database. By default, the new name is DEFAULT, so you're strongly encouraged to set this value. This option changes the initial size of one specific data dictionary index. A value of 30 makes it three times larger, for example. The default value (15) should be adequate for most users. By setting this option, you can change the National Language Standard (NLS) NCHAR character set used for your database. Not setting this option leaves your Oracle7 character set in place. This is the name of the parameter file to be used by the instance in which the migration will occur. Not setting this option causes the default file to be used. This option names the full path and filename where the Migration utility will write its log file. When the Migration utility completes its processing, you should check the spool file to see if any errors occurred.

DBNAME NEW_DBNAME

MULTIPLIER

NLS_CHAR

PFILE

SPOOL

Don't open the database as an Oracle7 database at this point; further conversion steps need to be completed before the database is usable again. Prematurely opening the database corrupts this intermediate version and you won't be able to complete the migration process successfully. Time to take a backup You should make a backup of this version of the database because it can be used as your first Oracle8 backup, as well as an intermediate starting point for another migration attempt. Moving or Copying the Convert File The Migration utility created a convert file for you in the Oracle7 environment. This file will be found in the DBS, or related directory, under the Oracle7 home directory and will be named CONVSID.DBF (where SID is the Oracle7 instance name). You'll need to move this file to the corresponding directory in the Oracle8 home directory, renaming it to reflect the Oracle8 instance name if this is different. If you aren't going to uninstall Oracle7 at this time, you can wait and complete the file transfer in a single step. If you're going to uninstall Oracle7, make a copy of this file outside the Oracle directory structure so that you can find it later. Installing the Oracle8 Version of Oracle If you don't have space for the Oracle8 installation, you can remove the Oracle7 directory structure before beginning this step. However, it is recommended that you back it up first, in case you need to use your Oracle7 database again. Use the Oracle7 installer to uninstall Oracle7 and the Oracle8 installer to add the Oracle8 files. Your platform-specific documentation explains how to run the installer for both operations. When installing Oracle8, be sure to select the Install/Upgrade option in order to prevent Oracle from creating a brand-new database that you won't need. Adjusting Your Environment Variables and Parameter File
http://www.informit.com/content/0789716534/element_003.shtml (12 of 16) [26.05.2000 16:46:27]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

You need to ensure that your operating system is aware of and using the new Oracle8 code before continuing with the migration process. The remaining migration tasks require the Oracle8 executables to manipulate your database. This means resetting the pointers to Oracle Home and related structures, whatever they might be for your operating system. Again, you need to refer to your platform-specific documentation if you aren't sure what these are. You also need to check your Oracle7 parameter file for obsolete or changed parameters. These are listed in the Oracle8 Server Migration Manual, available as part of the Oracle8 distribution media. Table 3.4 lists the non-platform-specific parameters that you need to address. Table 3.4 Obsolete and changed parameters Oracle7 Name: INIT_SQL_FILES LM_DOMAINS LM_NON_FAULT_TOLERANT PARALLEL_DEFAULT_SCANSIZE SEQUENCE_CACHE_HASH_BUCKETS SERIALIZABLE SESSION_CACHED_CURSORS SNAPSHOT_REFRESH_INTERVAL SNAPSHOT_REFRESH_PROCESS Obsolete: Yes Yes Yes Yes Yes Yes Yes No No Oracle8 Name:

JOB_QUEUE_INTERVAL JOB_QUEUE_PROCESSES

Use your favorite editor to make any necessary changes to your parameter file. You may also want to move it to a new directory so that it stays with your other Oracle8 files. If you use the default conventions for your parameter filename and location, see the Oracle8 documentation for your specific system to identify what these need to be. Removing or Renaming the Current Control and Convert Files You'll perform one conversion step a little later that will create new control files for your database. At this time, therefore, you should remove the control files your database was using. Drop them (if they're safely backed up) or rename them so that you can find them again if needed. If you've already uninstalled Oracle7, you should have copied the convert file to a safe place as discussed earlier in "Moving or Copying the Convert File." You should now move this copy to the appropriate directory in your Oracle8 Home directory structure. If you haven't uninstalled Oracle7, simply copy the file, renaming it if necessary, to the corresponding directory under Oracle8; see the earlier section titled "Moving or Copying the Convert File" for details. Starting an Instance Use Server Manager and the INTERNAL user to start an instance. You should then start a spool file to track the remaining conversion tasks performed on the database. You can use the following script to complete these steps by using Server Manager running in line mode: CONNECT INTERNAL STARTUP NOMOUNT SPOOL convert Complete the remaining database conversion activities 1. Issue the command ALTER DATABASE CONVERT to build new control files and update the data file header information. This is a point of no return! After ALTER DATABASE CONVERT completes, you can no longer use your database with Oracle7 code or programs. 2. Open the database, which will convert the rollback segments to their Oracle8 format. 3. Run the CAT8000.SQL script to do the following: Locating the CAT8000.SQL script and the log file

http://www.informit.com/content/0789716534/element_003.shtml (13 of 16) [26.05.2000 16:46:27]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

If you aren't in the directory where the CAT8000.SQL script is located, you need to include the full path name. You'll find this script in the Oracle home directory, under the ADMIN directory, which is under the RDBMS directory. After issuing the host prompt to check for the session's log, you should find the log in your current directory. It will be named CONVERT.LST, but the name may be case sensitive on some operating systems. Convert the rollback segments to their Oracle8 format. r Update data dictionary components. r Drop the MIGRATE user with the Oracle7 data dictionary. 4. Shut down the database if the preceding tasks are all successful. 5. Complete the post-migration tasks.
r

Perform these steps while still connected to your Server Manager session by using the following commands: ALTER DATABASE CONVERT; ALTER DATABASE OPEN RESETLOGS; @CAT8000.SQL HOST

SHUTDOWN Start the CONVERT.LOG file here to check for errors, and then EXIT back to Server Manager. If you find errors in the log file, you may need to repeat the tasks discussed in this section; you may instead, depending on the severity of the problem, have to repeat most or all of the migration process after correcting the cause of the errors. If you've completed your migration at this point, you can skip the following discussion of alternate migration techniques. Continue with the section "Completing Post-Migration Steps" to learn how to make your Oracle8 database available to your applications and users.

Executing the Migration with Export/Import or Table Copying

In this section you look at two other options for migrating your database that you may want to use instead of the Migration utility. Both require you to move data between an Oracle7 and Oracle8 database. Therefore, unlike the Migration utility process, you have to build an Oracle8 database yourself, or let the installer build one for you. In all probability, you will need to customize any database you build to match the structure of your Oracle7 database. The main difference between these two approaches is that the export/import option lets you work on each database independently, so you can remove your Oracle7 database before building and populating your Oracle8 database. This can be useful if space is at a premium. The table-copying method requires that the Oracle7 and the Oracle8 databases be online during the migration process. If you have the space, you can also leave both databases in place during an export/import. However, if the Oracle7 database is available to users following the export, the changed data will have to be identified and migrated separately. In the following sections, it's assumed that you're going to keep your Oracle7 database in place during the whole migration process, but it's pointed out when you could drop it depending on what method you're using. The following steps are based on this assumption. Step 1: Install Oracle8 Install Oracle8 on your system by using the Oracle installer as described in your platform-specific documentation. You can choose to have the installer build your initial Oracle8 database if you prefer. Step 2: Prepare Your Oracle8 Database If you've let the installer build your database, you simply need to add the tablespaces that match your current Oracle7 structure. You should also add the same number of rollback segments and redo logs as you're now using in Oracle7. If you're using the table-copying method, you also need to create the users at this point. Copying tables across database links

http://www.informit.com/content/0789716534/element_003.shtml (14 of 16) [26.05.2000 16:46:27]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

If you plan to use CREATE TABLE...AS SELECT commands to make table copies, you also need to build database links that allow the Oracle7 and Oracle8 databases to work together. Data-base links themselves are described in the Oracle8 SQL manual and in the Distributed Database documentation. If you aren't familiar with distributed processing and database links, this is probably not a good method to use for your migration. Step 3: Prepare to Migrate If you're performing the export/import process, you should now create the export file of your full database, or whatever pieces of the database you want to migrate. After this, you can shut down your Oracle7 database and uninstall Oracle7 if you want. If you're performing table copying, you need to define the network protocol and addresses for SQL*Net or Net8. SEE ALSO If you don't already have these tools configured, you might as well use Net8, which is discussed on Step 4: Move the Data Now you can move the data into the Oracle8 database. By using export/import, you simply execute the Oracle8 import command and provide the name of the file you exported in step 3. If you're performing table copying, you can use either the COPY command available in SQL*Plus or the SQL CREATE TABLE...AS SELECT command. The former identifies the target (Oracle8) or the source (Oracle7) database, or both, using SQL*Net or Net8 aliases from the TNSNAMES.ORA file. The latter uses a database link name in the new table name or the name of the table being copied, depending on where the command is running. If you're in the Oracle7 database, the link name is appended to the new table name; if you're in Oracle8, the link name goes on the source table name. Your database should be ready-if you used export/import-after you complete the data transfer. If you performed table copying, you may still need to duplicate the other objects in your Oracle7 database, such as indexes, views, synonyms, and privileges. The simplest way to do this is with an export/import of the full database. In this case, though, you wouldn't export the table rows and would have to allow the import to ignore errors due to existing tables.

Completing Post-Migration Steps

The following sections cover the steps needed to make the database accessible by the applications and the users of those applications. You might not need to follow each step exactly, depending on your system and application mix.

Precompiler Applications
Even if you don't intend to make any changes to your precompiler applications, you need to relink the applications before they will run against the Oracle8 database. You should relink them to the SQLLIB runtime library provided with the Oracle8 precompiler. Of course, if you want to take advantage of some new features of Oracle8, you need to modify your code and observe the standard precompile and compile steps.

OCI Applications
You can use your Oracle7 OCI applications with Oracle8 unchanged. If you have constraints in your applications, however, you should relink the applications with the Oracle8 runtime OCI library, OCILIB. You can choose a non-deferred mode to relink, in which case you'll experience Oracle7 performance levels, or you can use deferred mode linking to improve performance. The latter may not report any linking, bind, and define errors until later in the execution of the statements than you're used to seeing. Specifically, they will occur during DESCRIBE, EXECUTE, or FETCH calls rather than immediately after the bind and define operations. Obsolete OCI calls Two calls used in OCI programs, ORLON and OLON, are no longer supported in Oracle8; you should use OLOG in their place. Although OLOG was originally introduced for multithreaded applications, it's now required for single-threaded code.

http://www.informit.com/content/0789716534/element_003.shtml (15 of 16) [26.05.2000 16:46:27]

informit.com -- Your Brain is Hungry. InformIT - Migrating an Oracle7 Database to Oracle8 From: Using Oracle8

SQL*Plus Scripts
Ensure that your SQL*Plus scripts don't contain a SET COMPATIBILITY V7 command. If they do, change it to SET COMPATIBILITY V8. Also remember to check any LOGIN.SQL scripts for this command.

SQL*Net
The only severe problem you might run into with SQL*Net is if you're still using version 1. Oracle8 will only communicate via SQL*Net version 2 or Net8. The SQL*Net v2.0 Administrator's Guide and SQL*Net version 2 Migration Guide explain how to upgrade to version 2. As with other Oracle8 products, Net8 gives you a lot of additional features that you may want to consider using.

Enterprise Backup Utility (EBU)

Oracle8 has replaced the Enterprise Backup utility (EBU) with Recovery Manager (RMAN). Therefore, any code and routines you've developed around EBU will need to be replaced. In addition, the backup volumes created under EBU aren't usable by Oracle7. EBU and RMAN both use the same Media Management Language to talk to third-party storage subsystems, so you should still be able to use any tape subsystems and tape management modules that you used with EBU when you convert your backup routines to RMAN.

Standby Databases
A standby database must run on the exact same release as the production database that it mirrors. Therefore, you need to upgrade any standby database after you upgrade your Oracle7 production database. Migrate your standby database to Oracle8 1. Apply all redo logs created under Oracle7. 2. Ensure that the primary database is successfully opened under Oracle8. 3. Install Oracle8 on the standby database platform. 4. Copy the production database's control file and first data file to the standby site. 5. Make a new control file for the standby database. Impact of using new Oracle8 features If you begin using Oracle8's new features, you may have to make further changes to applications by using the products already discussed, and you may have to change code and procedures related to the tools listed here. For example, you have to run a CATEXP7.SQL script if you want to export Oracle8- partitioned tables to an Oracle7 database.

Migration: Final Considerations

The following Oracle products will run unchanged against your Oracle8 database: q Forms q Developer/2000 applications q PL/SQL q Export/Import You should consider the possible improvements you might obtain, however, if you begin using some of the appropriate Oracle8 enhancements. This doesn't have to be done immediately, of course, but over a period of weeks or months, as time permits. You should also ensure that the application developers are aware of the possible enhancements to their code. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_003.shtml (16 of 16) [26.05.2000 16:46:27]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Managing with Oracle Enterprise Manager (OEM)

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

Introducing OEM Components

r r r r

OEM Console Common Services Intelligent Agents Application Programming Interface (API) Minimum Requirements Compatibility Issues Performing the OEM Installation User and Repository Setup Starting the Intelligent Agent and the Listener Testing the Configuration

Installing and Configuring OEM

r r r r r r

q q

Setting Up Preferred Credentials Setting Up Security

r r

Examples of Client Files Required by Enterprise Manager Examples of Server Files Required by Enterprise Manager Starting and Stopping Your Database Managing Users and Privileges Using OEM's Navigator Menu to Manipulate Users and Privileges Managing Database Storage Using Oracle Performance Manager Using Oracle Expert Using Oracle TopSessions

Basic Management Tasks with OEM

r r r r

Performing Advanced Management Tasks with the Performance Pack

r r r

q q q

Install and configure Oracle Enterprise Manager Set up the Repository Manage users and privileges

http://www.informit.com/content/0789716534/element_004.shtml (1 of 20) [26.05.2000 16:46:43]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8
q q

Manage storage Tune your database

Introducing OEM Components

The Oracle Enterprise Manager combines a graphical console, agent processes, and common services to provide an integrated and comprehensive systems management platform for managing Oracle databases on the network. You can perform the following tasks from Enterprise Manager: Don't install OEM 1.2.2 and OEM 1.5.0 in Oracle 8.0.3 home OEM v1.2.2 isn't compatible with Oracle Server 7.3.3 and 8.0.x. OEM v1.5.0 shouldn't be installed in an Oracle Server 8.0.3 home directory. OEM's latest version (1.5.5), however, works fine for Oracle 7.3.3 and 8.0.x.
q q q q

Administer, diagnose, and tune multiple databases. Schedule jobs such as executing a SQL*Plus script on multiple nodes. Monitor objects and events such as database and node failures throughout the network. Integrate third-party tools.

Table 4.1 describes OEM's database application tools that allow you to perform the primary database administration tasks; Figure 4.1 shows these components. Figure 4.1 : OEM comprises various components that can be used for specific tasks. Backup Manager Data Manager Instance Manager Schema Manager Security Manager SQL Worksheet Storage Manager Lock Manager TopSessions Performance Manager Oracle Trace Navigator window Job Scheduling window Event Management window

Table 4.1 OEM components and their functions OEM Component Instance Manager TableSpace Manager Storage Manager Security Manager Schema Manager Server Manager Software Manager Backup Manager Data Manager SEE ALSO Using the Instance Manager, The Enterprise Manager environment consists of the following major components: q Oracle Enterprise Manager Console
http://www.informit.com/content/0789716534/element_004.shtml (2 of 20) [26.05.2000 16:46:43]

Function Manage instances, INIT.ORA file initialization parameters, and sessions Manage fragmentation and free space in tablespaces Manage tablespaces, data files, and rollback segments Manage users, roles, privileges, and profiles Manage schema objects such as tables, indexes, views, clusters, synonyms, and sequences Perform line-mode database operations from the client Manage the software distribution process Perform database backups and create backup scripts Perform export/import and data loads

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8
q

q q q q q

Services common to all OEM components, such as the Repository, discovery service, communication daemon, and job scheduling and event management systems Intelligent agent Integrated applications Application programming interfaces (APIs) Command Line Interface (CLI) Online help system

The basic OEM functionality is available to you with Oracle Server; however, you can install several optional management packs: Change Management Pack, Diagnostic Pack, and Tuning Pack.

OEM Console
The console user interface contains a set of windows that provide various views of the system. There's only one console per client machine. Table 4.2 describes the various components of the console. Table 4.2 Components of the console Component Navigator window Map window Job window Event Management window Description A tree view of all the objects in the system and their relationships Allows customization of the system views User interface to the Job Scheduling system User interface to the event management system

Common Services
The following services are common to various OEM components (see Table 4.3 for details on how these components interact): q The Repository is a set of tables in your schema, which can be placed in any Oracle database in the system. Information stored in the Repository includes the status of jobs and events, discovery cache, tasks performed, and messages from the notification queue in the communication daemon. To set up the Repository and manipulate it, you need to log on with an account that has DBA privileges. When logging in to Enterprise Manager, you're establishing a database connection into your Repository. At any given time, each user is connected to a single Repository. The connection to the Repository must be active during your working sessions. There can be more than one repository in the system. Repository tables can be installed on any Oracle database accessible to the console. A repository can be moved to another Oracle database. The Repository can be started or shut down from the Instance Manager but not from the console. Multiple repositories can exist within the same database You can use one repository, or you can switch between multiple repositories stored in the same database.
q

The communication daemon is a multithreaded process that manages console communication activities. It's responsible for communicating with agents and nodes for job scheduling and event monitoring, queuing and retrying failed jobs periodically, service discovery, contacting nodes periodically to determine their status, and maintaining a cache of connections to agents on nodes. The Job Scheduling System enables you to schedule jobs on remote sites by specifying the task to perform, the start time, and the frequency of execution. The Job System isn't usable if Oracle Intelligent Agent isn't installed and configured.

Reactive management is provided by the job and event systems You can use the Job and Event systems together to provide a reactive management system. This is achieved by allowing certain jobs to get executed when the specified events occur.
q

The Event Management System monitors events at remote sites, alerts you when a problem is detected, and optionally fixes it. In Windows NT, the application event log contains many of the errors detected by OEM. Security Services manages administrative privileges for nodes and services in the system. It manages a list of administrators who are notified when an event occurs. Service Discovery maintains an up-to-date view of the nodes and services being managed. The console's Navigator tree is populated with this information.

http://www.informit.com/content/0789716534/element_004.shtml (3 of 20) [26.05.2000 16:46:43]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

Table 4.3 Communication between OEM components Communication Path Console and communication daemon Description The console sends job and event requests to the communication daemon, and the status of these jobs and events are sent back to the console. Authentication requests of users logging in to the console are sent to the daemon. The daemon sends information to update the tree of nodes and services in the Navigator. Communication daemon and Job and event requests are handed to the Job or Event Management Common Services systems. The Common Services passes job and event status back to the communication daemon. Service Discovery information is passed from the Common Services to the daemon. Communication daemon and Agents communicate with the daemon to report results and status intelligent agent messages for jobs and events from the remote nodes. Common Services and Repository The Event Management and Job Management systems write event and job information, respectively, to the Repository. Figure 4.2 represents the communication path between different components of the Enterprise Manager in terms of the jobs, events, or any other requests logged in to the console. Figure 4.2 : Interaction between the various OEM components is well-defined.

Intelligent Agents
Intelligent agents are intelligent processes running on remote nodes. Each agent resides on the same node as the service it supports and can support all the services on that node. Intelligent agents perform the following functions: Use an intelligent agent to manage an older Oracle release Each intelligent agent is compatible with the database with which it's released and prior database releases. When used to manage an older release of the database, the intelligent agent must be installed in an ORACLE_HOME directory current with the agent release. Older releases of the intelligent agent aren't compatible with newer releases of the database.
q

q q q q q q q

Execute jobs or events from the console or third-party applications can be sent to the intelligent agent for execution. Cancel jobs or events as directed. Run jobs, collecting results and queuing them for the communication daemon. Run autonomously without requiring the console or the daemon to be running. Autonomously detect and take reactive measures (as specified by the administrator) to fix problems. Autonomously perform specified administrative tasks. Check and report events to the communication daemon. Handle SNMP requests if supported on the agent's platform.

An agent is required for all or some functionality of these components: Service Discovery; Job Control System; Event Management System; Backup Manager; Software Manager; Data Manager's Export, Import, and Load applications; Oracle Events; and Trace.

Application Programming Interface (API)

The APIs available with Enterprise Manager enable third-party applications-for example, applications that can analyze the data collected through Oracle Expert-to integrate the console with the Common Services. Third-party applications written in C++ that use OLE technology work very well with these APIs. Applications can be integrated at the console, service, or agent level; however, this integration depends on the third-party applications.

Installing and Configuring OEM

Several issues are involved in the installation and configuration of Enterprise Manager that you should address for the components to work together. Some issues include setting up the client, the server, and the Repository. Configuration involves setting preferred credentials and setting up security, among other things.
http://www.informit.com/content/0789716534/element_004.shtml (4 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

Not available for UNIX OEM is available only for Windows NT and Windows 95. However, the intelligent agent can run on UNIX or Windows NT.

Minimum Requirements
You need the following minimum hardware resources to install and use the OEM components: q Intel 486 PC or higher q VGA video (SVGA strongly recommended) q 32MB RAM q CD-ROM drive q Windows 95/NT-compatible network adapter q 25MB of hard disk space for Oracle Enterprise Manager, Net8, and required Oracle support files q 4MB of hard disk space for Oracle Enterprise Manager online documentation q 15MB of hard disk space for OEM's Performance Pack q Disk space for installing a local Oracle database or intelligent agent for Windows NT Installing documentation is optional The OEM documentation can take a lot of space. If you don't have enough disk space, you can run it from the CD-ROM when needed. The following minimum software resources are needed: q Microsoft Windows NT version 3.51 or higher, or Windows 95 q TCP/IP services

Compatibility Issues
Table 4.4 lists the components of Oracle Enterprise Manager version 1.5.0 and their compatibility with specific releases of Oracle Server. Table 4.4 Compatibility matrix for OEM 1.5.0 Feature Repository Local Remote Service Discovery Job Control System Event Management System Database Applications Backup Manager Instance Manager Schema Manager Security Manager Storage Manager SQL Worksheet Software Manager Utility Applications Data Manager/Export Data Manager/Import Data Manager/Load Performance Pack Expert Oracle Server 7.2 no yes yes yes yes yes yes yes yes yes yes no no no no yes Oracle Server 7.3 yes yes yes yes yes yes yes yes yes yes yes See /2/ yes yes yes yes Oracle Oracle Server 8.0.3 Server 8.0.4 See /1/ yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes

http://www.informit.com/content/0789716534/element_004.shtml (5 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

Lock Manager Oracle Events Performance Manager Tablespace Manager Top Sessions Trace /1/ /2/

yes no yes no yes no

yes yes yes yes yes yes

yes yes yes yes yes yes

yes yes yes yes yes yes

OEM 1.5 must be installed in a different home if there is a local 8.0.3 database. Software Manager can support Oracle Server 7.3.3 agents (Windows NT only) with upgraded OSM job files.

Performing the OEM Installation

Install and configure Enterprise Manager (general steps) 1. Configure Net8 locally and on the server. 2. Set up the client. 3. Set up the server. 4. Set up the Repository. 5. Install the OEM software on the client. 6. Install the intelligent agent on the server. The agent and the database that it services must be installed on the same node. Configuring Net8 You can use the following tools to generate the different files required for Net8 and Enterprise Manager: q Network Manager 3.1 for Windows to generate Net8 files and TNSNAMES.ORA q Oracle Topology Generator to generate TOPOLOGY.ORA q ASCII editor Net8 must be installed before installing OEM If Net8 isn't installed on the machine, select it from the OEM installer. You also can choose to install Performance Pack at this point. You need to configure Net8 so that it can use the database you want to access with OEM. You use Network Manager and Topology Generator to generate the TNSNAMES.ORA file, which will contain the information for the sample database. However, you need to edit this file with the Net8 easy configuration utility or a text editor so it will have information of other databases that you will use from OEM. Installing and Configuring the Intelligent Agent You can choose to install the intelligent agent as part of the Oracle Server installation, or you can install it later by running the Oracle installer. The following is required for the agent to function correctly: q Sun SPARCstation q Solaris version 2.4 q 32MB RAM q Oracle Server version 7.3 or higher q SQL*Net version 2 or higher (Net8) Before the agent is started, you must do the following to create a user account with appropriate privileges for the intelligent agent: q Run the CATSNMP.SQL script (found in the $ORACLE_HOME/rdbms/admin directory) from Server Manager to create a user for the intelligent agent and give that user DBA privileges. SEE ALSO Creating a user, Granting privileges to roles, q The SNMP.ORA file must be edited with any text editor. This file contains the address of the agent and

http://www.informit.com/content/0789716534/element_004.shtml (6 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

other listener information, which is read by the agent. It resides in the $ORACLE_HOME/network/admin directory on the database server. Install Oracle Enterprise Manager 1. Log in to Windows NT as the administrator or a user with permissions equivalent to an administrator. 2. Change to the \NT_x86\INSTALL directory on the CD-ROM drive. 3. Double-click ORAINST.EXE or SETUP.EXE to launch the Oracle installer. 4. Select Oracle Enterprise Manager to install the base product. The Installer will search for the TOPOLOGY.ORA and TNSNAMES.ORA files in the ORACLE_HOME\network\admin directory. If TOPOLOGY.ORA isn't found, an error message appears. If TNSNAMES.ORA is found but not TOPOLOGY.ORA, you'll be prompted to create the TOPOLOGY.ORA file by using the Oracle Network Topology Generator. If the TNSNAMES.ORA file isn't found, you can use the Oracle Network Manager to create the file. 5. Exit the installer after installation is complete. 6. Log off from Windows NT and then log in again. 7. If a local Oracle NT database is being accessed, you need to use Control Panel's Services tool to verify that the Oracle Service is started, and then start up the local NT database.

User and Repository Setup

Before Oracle Enterprise Manager is used, you must create a set of base tables that contain environment information for the managed databases-this is the Repository. You create the necessary tables in the Repository by using the SMPCRE.SQL and XPOCR.SQL scripts found in the $ORACLE_HOME/ rdbms/admin directory. An Oracle user must be created with appropriate permissions to access the Repository before the scripts are run. For each user that needs to access the console, a separate Repository must be created and setup scripts must be run. Console and repository compatibility The Repository must be compatible with the version of the Oracle Enterprise Manager. If the Repository version is older or newer than the console version, you must install a more recent compatible version of Enterprise Manager. Set up the user and Repository 1. Create a new user with Server Manager: SVRMGR> create user sysman identified by sysman 2. Grant this user the same privileges as SYSTEM: SVRMGR>grant dba to sysman 3. Connect as the new user: SVRMGR> connect sysman/sysman@testdb 4. Execute SMPCRE.SQL and XPOCR.SQL to build the tables and views required by the console and Expert: SVRMRG>@smpcre.sql SVRMGR>@xpocr.sql The user sysman can now log in to the Oracle Enterprise Manager.

Starting the Intelligent Agent and the Listener

For Enterprise Manager to connect and work successfully, the intelligent agent and the listener must be started on the server. For the client to communicate with the server, the communication daemon must be running. Starting the daemon The communication daemon is started and shut down automatically when Enterprise Manager is started. The following shows how to run the agent and the listener:
http://www.informit.com/content/0789716534/element_004.shtml (7 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

Task Start the agent Shut down the agent View agent's status Start the listener on UNIX Shut down the listener on UNIX

Command c: > net start oracleagent c: > net stop oracleagent c: > net start $ lsnrctl start testdblsnr $ lsnrctl stop testdblsnr

Start/stop the listener in Windows NT Use the Control Panel's Services tool to start and stop the listener in Windows NT.

Testing the Configuration

Test the configuration 1. Shut down the listener and the agent. $ lsnrctl stop testlsnr $ lsnrctl dbsnmp_stop 2. Start up and log in to Enterprise Manager. You should get the message ORA-12224: TNS: no listener occurs, and login is not possible. Enterprise Manager requires that the listener run on the server at all times. Logging in to OEM doesn't require the agent to be running Log in is possible without the agent running, because the agent is required only for jobs and events submitted or returned by the remote database or the console. 3. Start up the listener: $ lsnrctl start testlsnr 4. Double-click the Oracle Enterprise Manager icon from Program Manager and log into the database as sysman connected as sysdba (see Figure 4.3). Figure 4.3 : Log in to OEM by providing the information requested. The username to use for connecting to the database (for example, sysman) The password for the username The Net8 service name for the database to which you're connecting Connect as Normal, SYSOPER, or SYSDBA

Setting Up Preferred Credentials

You set up preferred credentials to avoid retying the service name, service type, and username for each database, listener, or node that the user intends to access. The Preferred Credentials page of the User Preferences dialog box shows the list of databases, listeners, and nodes in the network (see Figure 4.4). Figure 4.4: You can set preferred credentials from the console. Set the preferred credentials 1. From the console's File menu, choose Preferences. 2. Select any service from the list of entries in the dialog box and populate the Username, Password, Confirm, and Role text boxes.

Setting Up Security
The following operations on a remote instance require that security be set up for Enterprise Manager users: q STARTUP q SHUTDOWN
http://www.informit.com/content/0789716534/element_004.shtml (8 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8
q q q q q q q q

ALTER DATABASE OPEN and MOUNT ALTER DATABASE BACKUP ARCHIVE LOG RECOVER All system privileges with ADMIN OPTION CREATE DATABASE Time-based recovery

Set up remote security 1. Create a password file by connecting to the oracle user account, changing to the ORACLE_HOME/dbs directory, and then using the orapwd utility in UNIX: $ orapwd file=orapwtestdb password=testpass entries=10 In this example, the SID is assumed to be testdb. 2. Grant appropriate roles: SVRMGR> grant sysdba to sysman SVRMGR> grant sysoper to sysman 3. Edit the INIT.ORA file to add the following entry: REMOTE_LOGIN_PASSWORDFILE=EXCLUSIVE 4. Shut down the instance. At this point, the database instance can be shut down from Enterprise Manager, but local security needs to be set up on Windows NT clients to start up the database from OEM. Set up local security 1. Download the INIT.ORA and CONFIG.ORA files from the server and copy them into the \OEM_directory\dbs directory on the Windows NT client. 2. On the client, edit the INIT.ORA file by using any text editor like Notepad, and change the "ifile" entry to the directory in which the CONFIG.ORA file is located, with the "ifile" set to the CONFIG.ORA file. 3. Restart the Enterprise Manager.

Examples of Client Files Required by Enterprise Manager

OEM uses several files on the client side-SQLNET.ORA, LISTENER.ORA, and TOPOLOGY.ORA-in the $ORACLE_HOME/network/admin directory. SQLNET.ORA contains optional parameters that can be used for tracing Net8 connections, whereas LISTENER.ORA is used to provide the address on which the listener listens. SQLNET.ORA ===========================================================

===========================================================

http://www.informit.com/content/0789716534/element_004.shtml (9 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

ORACLE_HOME is set to c:\orant Because the default domain and zone are set to world, the service names in TNSNAMES.ORA should have world tagged to them TNSNAMES.ORA ===========================================================

=========================================================== The domain world must match the SQLNET.ORA file Should match the port in the SNMP.ORA file Should match the port in the LISTENER.ORA file The database and SID name is test The port numbers in TNSNAMES.ORA must be unused by any other service and must be valid port numbers as per TCP/IP standards. TOPOLOGY.ORA ===========================================================
http://www.informit.com/content/0789716534/element_004.shtml (10 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

=========================================================== Should match agent name in TNSNAMES.ORA Should match listener name in LISTENER.ORA Database name should match the one in TNSNAMES.ORA

Examples of Server Files Required by Enterprise Manager

The agent uses several files on the server side-SQLNET.ORA, TNSNAMES.ORA, and LISTENER.ORA-that reside in the $ORACLE_HOME/network/agent directory. SQLNET.ORA contains optional parameters that can be used for tracing Net8 connections, whereas LISTENER.ORA is used to provide the address on which the listener listens. SQLNET.ORA =========================================================== AUTOMATIC_IPC=OFF trace_level_server=off TRACE_LEVEL_CLIENT=off SQLNET.EXPIRE_TIME=0 NAMES.DEFAULT_DOMAIN=world NAME.DEFAULT_ZONE=world SQLNET.CRYPTO_SEED="-2089208790-14606653312" =========================================================== TNSNAMES.ORA ======================= ===================================== #Database Addresses test.world=(description= (address_list= (address= (community=tcp.world) (protocol=tcp) (host=fastmachine) (port=1701) ) ) (connect_data= (sid=test) (global_name=test.world) ) ) =========================================================== LISTENER.ORA ===========================================================

http://www.informit.com/content/0789716534/element_004.shtml (11 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

=========================================================== Listener name, domain, host name, and SID are the same in all the other files Listener name, domain, host name, and SID are the same in all the other files Must match the port in TNSNAMES.ORA on the client and server machines Listener name, domain, host name, and SID are the same in all the other files SNMP.ORA ===========================================================

http://www.informit.com/content/0789716534/element_004.shtml (12 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

=========================================================== Listener name, SID, and host name are the same in the other files Must match exactly with agent address in TNSNAMES.ORA on client machine

Basic Management Tasks with OEM

As an Oracle DBA, you'll be performing several tasks on a daily basis on OEM, such as starting up shutting down the database and managing users.

Starting and Stopping Your Database

After you set up remote and local security, you can start or shut down an Oracle database from the Enterprise Manager console. To start up or shut down a database, you must have the SYSOPER or SYSDBA role. Start up the database 1. Start up and log in to the Enterprise Manager. 2. Double-click the database object to be started. If you haven't set up preferred credentials for this host, you'll get an error message. Type the correct username and password each time, or set up preferred credentials. 3. In the database property sheet that appears, choose the Startup option. 4. Click the appropriate startup option and specify the location of the INIT.ORA file. 5. Click the startup button. Shut down the database 1. Start up and log in to Enterprise Manager. 2. Double-click the database object to be stopped. If preferred credentials haven't been set up for this host, an error message occurs. Type the correct username and password each time, or set up preferred credentials. 3. In the database property sheet that appears, choose the Shutdown option. 4. Click the appropriate shutdown option and then click the shutdown button.

Managing Users and Privileges

You can easily manage users and privileges by using OEM's Security Manager component (see Figure 4.5). You can manage user information for several databases from one centralized location. Figure 4.5 : Security Manager enables you to manage users, profiles, and roles. After Security Manager successfully connects to the database, you see a tree structure with three context-sensitive objects. The database name is displayed next to the database container, and the Users, Roles,

http://www.informit.com/content/0789716534/element_004.shtml (13 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

and Profiles containers branch from the current database container. You can use Security Manager's User menu to create, edit, or remove existing users on a database. Manipulate roles and profiles from the menus Roles and profiles can also be similarly created, edited, and removed by using the Roles and Profiles menus. Create a user 1. From the User menu choose Create. 2. In the Create User property sheet, enter the new user's username in the Name text box (see Figure 4.6). Figure 4.6 : You can create users from the property sheet. 3. 4. 5. 6. In the Authentication section, enter a password and then re-enter to confirm it. Choose the appropriate default and temporary tablespaces for the user from the Tablespaces section. Click Create. Verify the user creation by checking the Users object in Security Manager's tree structure. This verification can also be done by logging in as the new user with the password.

Quick-edit a user 1. Right-click the username to be modified. 2. Select Quick Edit from the pop-up menu. 3. Make the desired changes to the quotas, privileges, and roles. 4. Click OK. Remove a user 1. Select the username to be removed. 2. From the User menu, choose Remove. 3. In the confirmation dialog box, click Yes. The user can also be removed by right-clicking the highlighted username and choosing Remove from the pop-up menu. The User menu can be used to give privileges to users. Assign privileges to users 1. From the User menu choose Add Privileges to Users. 2. In the Add Privileges to Users dialog box (see Figure 4.7), select the user to which privileges are to be granted. Ctrl+click additional users in the list to select more than one user. Figure 4.7 : Privileges can be assigned to users in the Add Privileges to Users dialog box. 3. Select the Privilege Type (Roles, System, or Object). 4. Select the privileges to be granted. Ctrl+click additional privileges in the list to select more than one privilege. 5. Click Apply. The Security Manager's Profile menu can be used to assign existing profiles to existing users. Assign profiles to users 1. From the Profile menu choose Assign Profile to Users. 2. In the Assign Profile dialog box (see Figure 4.8), select the user or users to whom profiles are to be assigned. Figure 4.8 : Users can be assigned profiles in the Assign Profile dialog box. 3. Select the profile to assign. 4. Click Apply. 5. Additional profiles can be assigned by repeating steps 3 and 4. Click OK when all profiles are assigned.

Using OEM's Navigator Menu to Manipulate Users and Privileges

Create a user with the Navigator menu 1. In OEM's navigator tree, click the + to the left of the Databases folder. 2. Click the + to the left of the database name.
http://www.informit.com/content/0789716534/element_004.shtml (14 of 20) [26.05.2000 16:46:44]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

3. Click the + to the left of the Users folder. 4. From the Navigator menu, select Create User. 5. Enter the User information and click OK. Copy a user between databases 1. In OEM's navigator tree, click the + to the left of the Databases folder. 2. Click the + to the left of the database name. 3. Click the + to the left of the Users folder. 4. Select the username to be copied. 5. Drag and drop the username from one database to the other database folder. Manage database user properties such as quotas, roles, and privileges 1. In the navigator tree, click the + to the left of the Database folder. 2. Click the + to the left of the database name. 3. Click the + to the left of the Users folder. 4. From the Navigator menu, select Alter User. 5. On any of the four tabbed pages (General, Quotas, Privileges, or Default Roles), select the desired types. 6. Click OK.

Managing Database Storage

You can use Storage Manager (see Figure 4.9) to perform administrative tasks associated with managing database storage, such as managing tablespaces and rollback segments and adding and renaming data files. Figure 4.9 : You can use Storage Manager to manipulate tablespaces, data files, and rollback segments. You can use Oracle's Tablespace Manager to monitor and manage database storage. It can be used to display graphically how storage has been allocated for the database segments, to defragment segments, and to coalesce free adjacent blocks. Monitoring Tablespaces Tablespace Manager's main window includes a tree list on the left and a drill-down on the right for a detailed view. You use the Tablespace Manager as follows: q Click the Tablespaces container to display the space usage for each tablespace in the instance. q Double-click the Tablespaces container for additional information about the tablespaces. q Click an individual tablespace container to display the Segments page for that tablespace. Clicking an individual segment graphically displays the space utilization for that segment. q Click a data file container to find the space usage for a data file. q Click an individual data file to display the Segments page for that data file (see Figure 4.10). Figure 4.10: Use Tablespace Manager to manage data files.

Performing Advanced Management Tasks with the Performance Pack

As a DBA, you should frequently monitor your system resources to identify contention. OEM provides various predefined charts that can help you in monitoring the usage of different resources that can contribute to contention. (For additional information on how to identify and reduce various types on contention, see Chapter 21, "Identifying and Reducing Contention.") Three resources need to be carefully monitored:
q

CPU. Every process that executes on the server needs some time slice of the CPU time to complete its task. Some processes need a lot of CPU time, whereas others don't. You should be able to identify your CPU-intensive processes.

My tuning philosophy Performance tuning shouldn't be treated as a reactive strategy; instead, it should be a preventive action based on trends detected through analysis by using tools such as the Performance Pack.
q

Disk access. Every time a process needs data, it will first look at the buffer cache to see if the data is

http://www.informit.com/content/0789716534/element_004.shtml (15 of 20) [26.05.2000 16:46:45]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

already brought in. If the data isn't found, the process will access the disk. Disk access is very time-consuming and should be minimized. Memory. Insufficient memory can lead to performance degradation. When the system falls short on memory, it will start paging and eventually start swapping physical processes.

The Performance Pack is a value-added component of the Oracle Enterprise Manager. It provides various tools to monitor and tune the performance of your database. It's important to understand that taking a point-in-time snapshot of the system doesn't do performance tuning, but it's a way to take into consideration the system performance over a period of time. You can perform three different types of tuning by using the Performance Pack components (see Table 4.5). Table 4.5 Types of tuning available through the Performance Pack Tuning Type Routine Tuning Focused Tuning What-If Tuning Description Used to identify and solve potential problems before they occur Used to resolve known performance problems Used to determine what would happen if a particular configuration change is made

The Performance Pack provides several tools (see Table 4.6) to capture, store, and analyze information so you can improve overall performance. Table 4.6 Performance Pack components and their functions Component Performance Manager Oracle Expert Function Displays tuning statistics on contention, database instance, I/O, load, and memory within predefined or customized charts Collects and analyzes performance-tuning data on predefined rules, generates tuning recommendations, and provides scripts that help with the implementation of tuning recommendations Collects performance data based on events and generates data for the Oracle Expert Displays the top 10 sessions based on any specified sort criteria Displays the free space left on each data file Displays the blocked and waiting sessions Monitors the specified conditions in the databases, nodes, and networks

Oracle Trace Oracle TopSessions Monitor Tablespace Viewer Oracle Lock Manager Oracle Advanced Events

To start the performance-monitoring applications from the OEM console, use the Performance Pack launch palette or the Performance Pack option on the Tools menu.

Using Oracle Performance Manager

Performance Manager is a tool for monitoring database performance in real-time. It provides a number of predefined charts for displaying various statistics in different formats, including tables, line charts, bar charts, cube charts, and pie charts (see Figure 4.11). Figure 4.11: Read consistency hit ratio is one type of information that can be charted. Performance Manager's Display menu includes items for seven different categories of predefined charts. Table 4.7 describes these categories and the set of charts that focus on displaying information of that category. Table 4.7 Charts used to identify contention Category Contention Database_Instance Charts Included in This Category Circuit, Dispatcher, Free List Hit %, Latch, Lock, Queue, Redo Allocation Hit %, Rollback NoWait Hit %, and Shared Server Process, Session, System Statistics, Table Access, Tablespace, Tablespace Free Space, #Users Active, #Users Waiting for Locks, and #Users Running File I/O Rate, File I/O Rate Details, Network I/O Rate, and System I/O Rate

I/O

http://www.informit.com/content/0789716534/element_004.shtml (16 of 20) [26.05.2000 16:46:45]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

Load Memory

Overview

User-Defined

Buffer Gets Rate, Network Bytes Rate, Redo Statistics Rate, Sort Rows Rate, Table Scan Rows Rate, and Throughput Rate Buffer Cache Hit %, Data Dictionary Cache Hit %, Library Cache Hit %, Library Cache Details, SQL Area, Memory Allocated, Memory Sort Hit %, Parse Ratio, and Read Consistency Hit % #Users Active, #Users Logged On, #Users Running, #Users Waiting, Buffer Cache Hit, Data Dictionary Cache Hit, File I/O Rate, Rollback NoWait Hit %, System I/O Rate, and Throughput Rate Charts created by the user

By default, information in the predefined charts is presented in the following manner: q Charts showing rates per unit of time are presented as line charts. q Charts showing ratios are presented as pie charts. q Charts consisting primarily of text information are presented as tables. q Charts displaying a large number of instances are presented as tables. The overview charts are a set of 12 predefined charts that give a good overall picture of the system (see Table 4.8). Table 4.8 Predefined charts Chart Number of Users Active Number of Users Logged On Description Shows the number of users actively using the database instance. Obtains information from the V$SESSION view. Shows the number of concurrent users logged on to the database instance, regardless of whether any activity is being performed. Obtains information from V$LICENSE. Shows the number of concurrent users logged on to the database instance and now running a transaction. Obtains information from V$SESSION_WAIT. Shows the number of users now waiting. Obtains information from V$SESSION_WAIT. Shows the buffer cache hit percentage. Obtains information from V$SYSSTAT. Shows the Data Dictionary cache hit. Obtains information from V$ROWCACHE. Shows the number of physical reads and writes per second for each file of the database instance. Obtains information from V$DBFILE. Shows the hits and misses for online rollback segments. Obtains information from V$ROLLSTAT. Shows I/O statistics including buffer gets, block changes, and physical reads per second for the database instance. Obtains information from V$SYSSTAT. Shows the number of user calls and transactions per second for the instance. Obtains information from V$SYSSTAT.

Number of Users Running

Number of Users Waiting Buffer Cache Hit % Data Dictionary Cache Hit File I/O Rate Rollback NoWait Hit % System I/O Rate

Throughput Rate

Get an overall picture of activity on a database with the Overview chart 1. In the navigator window, select the ORCL database and then click the Oracle Performance Manager icon. 2. From the Monitor menu, click Display and then choose Overview. Monitor disk access, resource contention, and memory utilization 1. Launch the Oracle Performance Manager in the context of the ORCL database, as explained in step 1 of the previous section. 2. From the Charts menu choose Define Window. 3. In the Window Name text box, provide a unique name. 4. Scroll through the list of available charts, select the chart you want, and click the << button. 5. Repeat step 4 for all the charts you need, and then click OK.
http://www.informit.com/content/0789716534/element_004.shtml (17 of 20) [26.05.2000 16:46:45]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

If the predefined charts don't suit your needs, you can create your own charts and save them for future use. Creating your own charts 1. From the Charts menu, choose Define Charts. 2. Click the New Chart button. 3. Enter a name for the new chart. 4. In the SQL Statement text box, enter a statement that will gather the statistics to display in the chart. 5. Click the Execute button. 6. Verify the results in the results field. 7. On the Display Options page, enter the required information for each variable you want to display and click the Add button. 8. Click the Apply button. 9. Click OK. 10. From the File menu choose Save Charts, and save the chart in the Repository. Recording Data for Playback You can choose to record data in a chart for analysis at a later time. The collection size varies based on the polling interval, database activity at the time, and the collection interval. Collect historical data 1. Display the charts from which you want to collect data. 2. From the Record menu choose Start Recording. 3. Provide a unique name in the Data Collection Name dialog box and click OK. 4. When finished with the data collection, choose Stop Recording from the Record menu. 5. Provide the database connect string in the Format/Playback Login dialog box. Playback recorded data 1. From the Record menu choose Playback. 2. In the Format/Playback Login dialog box, provide the connect string on the database where the formatted data is saved. 3. Select the data collection to play back and click OK.

Using Oracle Expert

Oracle Expert is a tool in the Performance Pack that you can use to tune a database. All tuning inputs and recommendations are stored in a tuning repository that allows the review and modification of the data and the rules at a later time. It has a knowledge base of tuning rules, designed through a tight relationship between the Oracle Server, Oracle Trace, and Oracle Expert development teams. It provides an explanation for all the recommendations it makes. Don't run multiple sessions of Oracle Expert against the same repository You can run multiple sessions of Oracle Expert against the same repository, but it's not recommended because it can lead to data conflicts between sessions. You should use Oracle Expert to complement your tuning experience, not as a tool to replace your function as a database performance tuner. You should instead focus on what to do with the findings and suggestions provided by Oracle Expert and enhance the rules used by Oracle Expert in analyzing the performance data. You can use Oracle Expert to tune the following: q Instance. It consists of tuning the SGA parameters (which affect the total size of the instance's system global area), I/O parameters (which affect the throughput or distribution of I/O for the instance), parallel query parameters, and sort parameters. q Application. It consists of tuning SQL and access methods, which involves determining the indexes needed and creating, modifying, and deleting them as needed. q Structure. It consists of sizing (recommendations for choosing storage parameters) and placement (recommendations on placing data files and partitioning segments). Increase the information analyzed by Oracle Expert

http://www.informit.com/content/0789716534/element_004.shtml (18 of 20) [26.05.2000 16:46:45]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

Before performing instance SGA tuning, run XPVIEW.SQL (in $ORACLE_HOME\rdbms\ admin) against the database being tuned to get better recommendations from Oracle Expert. Doing so causes Oracle Expert to collect additional information about the data-base's shared SQL segment. For Oracle Expert to perform data collection, the target database being tuned should have the following tables: dba_tab_columns, dba_constraints, dba_users, dba_data_files, dba_objects, dba_indexes, dba_segments, dba_ind_columns, dba_tables, dba_rollback_segs, dba_sequences, dba_views, dba_tablespaces, dba_synonyms, dba_ts_quotas, and dba_clusters. Oracle Expert doesn't collect information regarding index-only tables, partitioned tables, partitioned indexes, object types, object tables, and object views. Use Oracle Expert to gather tuning information (general steps) 1. Set the scope of the tuning session to tell Oracle Expert what aspects of the database to consider for tuning purposes. Oracle Expert collects the following categories of data: database, instance, schema, environment, and workload. 2. The collected data is organized in a hierarchical format. You can view and edit the rules and attributes used by Oracle Expert. 3. Oracle Expert generates tuning recommendations based on the collected and edited data. You can decide to use the recommendations, or ignore them and let Oracle Expert generate a new recommendation. 4. When you're satisfied with the recommendations, you can let Oracle Expert generate parameter files and scripts to implement the chosen recommendations. Don't tune the SYS or system schema Don't use Oracle Expert to tune the SYS or system schema. You should let Oracle tune these items automatically. Start an Expert Tuning session 1. From the File menu choose New. 2. Define the scope of the tuning session. 3. On the Collect page, specify the amount and type of data to collect. 4. Click the Collect button to acquire the required data. 5. On the View/Edit page are the rules used by Expert. You can modify the rules based on your experience. 6. On the Analyze page, click the Perform Analysis button to begin the data analysis. 7. Select Review Recommendations to review the recommendations provided by Expert. 8. If you agree with the recommendations, you can implement them by generating the requisite scripts and parameter files from the Implement page. If you don't agree with the recommendations, you'll have to change one or more rule and re-analyze (without recollecting) the data. Have enough privileges to perform some functions If the database management functions are grayed out from the menu bar, it may be because you aren't authorized to perform those functions. Reconnect as SYSOPER or SYSDBA. The collection classes to use are determined by the selected tuning categories for a tuning session. Reuse collected data When tuning multiple categories, the common classes need to be collected only once because Oracle Expert will be able to reuse the data for analysis. Start Oracle Expert 1. In the OEM map or navigator window, select a database and then click the Oracle Expert icon in the Performance Pack palette. Or double-click the Expert icon in OEM's Program Manager. 2. Connect to a tuning repository. 3. From the File menu choose New to create a new tuning session. 4. Enter the appropriate data in the dialog box pages. 5. Click OK. Permissions to use Oracle Expert The user running Oracle Expert must have SELECT ANY TABLE privilege for the database in which the repository is stored.
http://www.informit.com/content/0789716534/element_004.shtml (19 of 20) [26.05.2000 16:46:45]

informit.com -- Your Brain is Hungry. InformIT - Managing with Oracle Enterprise Manager (OEM) From: Using Oracle8

Using Oracle TopSessions

You can use the Oracle TopSessions utility to view the top Oracle sessions based on specified criteria, such as CPU usage and disk activity. Before running TopSessions for Oracle8, run $ORACLE_HOME/sysman/smptsi80.SQL to create all the supporting tables. Identify Oracle sessions that use the most CPU 1. In OEM's navigator window, select the ORCL database and then click the Oracle TopSessions icon. 2. On the Sort page of the Options property sheet (see Figure 4.12), change the Statistics Filter to User and the Sort Statistic to CPU Used by This Session. Figure 4.12: You can use the Sort page of the Oracle TopSessions Options dialog box to specify the criteria to use for monitoring sessions. 3. On the Refresh page of the Options property sheet (see Figure 4.13), select Automatic for the refresh type, set the Refresh Interval to 10, and reset the Minutes and Hours to 0. Figure 4.13: The Refresh page of the Options dialog box can be used to change the refresh type and refresh interval. 4. On the Count page of the Options property sheet (see Figure 4.14), select the Display Top N Sessions button and change the count to 10. Figure 4.14: The Count page of the Options dialog box can be used to specify the number of sessions to track. 5. Click OK to show the results (see Figure 4.15). Figure 4.15: Oracle TopSessions shows the results as specified by the resource usage criteria. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_004.shtml (20 of 20) [26.05.2000 16:46:45]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

You are here : Home : Using Oracle8

Managing Your Database Space

Exact Phrase All Words Search Tips

< Back

Contents

Next >

Save to MyInformIT

From: Using Oracle8 Author: David Austin Publisher: Que More Information

Space Management Fundamentals

r r r

Suggested Tablespaces Understanding File Types: File Systems Versus Raw Devices Understanding the Benefits of Striping Data Creating a Tablespace Setting Default Storage Values Changing the Characteristics of a Tablespace Dropping Tablespaces Comparing Dynamic and Manual Extent Allocation Releasing Unused Space Defragmenting Free Space

Adding New Tablespaces

r r

Tablespace Management
r r

Extent Allocation
r r r

q q q q q

Identify different segment types Design tablespaces for different segment types Manage tablespaces Make effective use of physical file structures Control unused and free space

Space Management Fundamentals

The basic storage unit in an Oracle database is the Oracle block (or database block). This is the smallest unit of storage that will be moved from disk to memory and back again. Oracle block sizes range from 2KB to 32KB. You obviously couldn't store a complete table on a single block, so blocks are grouped into segments to store amalgams of data. There are various types of segments: The DBA's role in space management Space management is an ongoing task for most of you. Unless you have a completely static database, tables and indexes will regularly grow, or shrink, in size. You need to ensure that sufficient space is available for this to occur without interruption to the ongoing processing. You also need to help ensure that the space is being used efficiently.
q

Data segments contain rows from a single table or from a set of clustered tables. SEE ALSO See how to use clusters on Index segments contain ordered index entries.

http://www.informit.com/content/0789716534/element_005.shtml (1 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8
q q q

LOB segments contain long objects. LOB index segments contain the special indexes for LOB segments. Rollback segments store "before" images of changes to data and index blocks, allowing the changes to be rolled back if needed. Temporary segments hold the intermediate results of sorts and related processing that are too large to be completed in the available memory. A single cache segment (also known as the bootstrap segment) holds boot information used by the database at startup.

You may have very large segments in your database, and it may be impossible to put the whole thing into a set of contiguous Oracle blocks. Oracle therefore builds its segments out of extents, which are sets of logically contiguous blocks. "Logically contiguous" means that the operating system and its storage subsystems will place the blocks in files or in raw partitions so that Oracle can find them by asking for the block offset address from the start of the file. For example, block 10 would begin at the (10XOracle block size) byte in the data file. It doesn't matter to Oracle if the operating system has striped the file so that this byte is on a completely different disk than the one immediately preceding it. The database always accesses blocks by their relative position in the file. A large segment may have several, or even several hundred, extents. In some cases it will be too big to fit into a single file. This is where the last Oracle storage construct plays a part. Rather than force users to deal with individual files, Oracle divides the database into logical units of space called tablespaces. A tablespace consists of at least one underlying operating system file (or database data file). Large tablespaces can be composed of two or more data files, up to 1,022 files. Every segment or partition of a partitioned segment must be entirely contained in a single tablespace. Every extent must fit entirely inside a single data file. However, many extents can comprise a partition or a non-partitioned segment, and the different extents don't all have to be in the same data file. Only one type of database object, a BFILE (binary file), is stored directly in an operating system file that's not part of a tablespace. There are six reasons to separate your database into different tablespaces: q To separate segments owned by SYS from other users' segments Recommendation for using multiple tablespaces Although Oracle doesn't prevent you from creating all your segments in a single tablespace, Oracle strongly recommends against it. You can control your space more easily by using different tablespaces than you can if everything were placed in a single tablespace. Multiple files let you build your tablespace in a manner that helps you improve space usage as well as database performance.
q q q q q

To manage the space available to different users and applications To separate segments that use extents of different sizes To separate segments with different extent allocation and deallocation rates To distribute segments across multiple physical storage devices To allow different backup and related management cycles and activity

The first reason to use multiple tablespaces is to keep the segments owned by user SYS away from any other segments. The only segments SYS should own-and the only ones that Oracle will create and manage for you-are those belonging to the data dictionary. The data dictionary The data dictionary is the road map to all the other database segments, as well as to the data files, the table-space definitions, the Oracle user-names, passwords and related information, and many other types of database objects. The dictionary needs to be modified as objects are added, dropped, or modified, and it must be available at all times. By keeping it in its own tablespace, you're less likely to run out of room (which would bring the database to a complete halt if it prevented SYS from modifying the dictionary). A second reason to use different tablespaces is to control how much space different schemas can take up with their segments. Each user can be assigned just so much space in any tablespace and doesn't need to be assigned any space at all in some tablespaces. Some end users may have no space allocated to them at all because their database access consists solely of manipulating segments that belong to the application owner. The third reason to manage your segments in different tablespaces has to do with space usage by the extents
http://www.informit.com/content/0789716534/element_005.shtml (2 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

belonging to different segments. As you can imagine, most databases have segments that are large and some that are small. To make effective use of the space, you would assign different-sized extents to these objects. If you mix these extents in a single tablespace, you may have problems if you need to drop or shrink the segments and try to reuse the freed space. Consider the example of a kitchen cabinet where you keep all your canned goods on one shelf (see Figure 5.1). After a trip to the grocery store, you can almost fill the shelf with cans of different sizes. Suppose that during the week you take out half a dozen or so of the smaller cans. After another trip to the grocery store, you couldn't simply put large cans in the spaces where the small ones came from, even if there were fewer large cans. You might have enough space if you rearrange the cans, but none of the spots is initially big enough for a single, large can. Figure 5.1 : You may have space management problems if you try storing different-sized objects together. Now suppose that you take out only large cans and then put medium-sized cans in their place. You still have some space around the medium-sized cans, but not enough to store even a small can. Again, you could find room for a small can, but only by shifting things around, as shown in Figure 5.2. Figure 5.2 : You'll run into space problems when replacing objects of different sizes. When large and small extents try to share limited tablespace storage space, they can "behave" like the cans. Space can be freed when extents are removed, but it's not necessarily of a useful size. Unlike you reorganizing the kitchen cupboard, however, the database can't easily shift the remaining extents around to make the space more useful. When space in a tablespace consists of an irregular checkerboard of used and free block groups, it's said to be fragmented. Much of the free space may never be reusable unless the extents are somehow reorganized. You can prevent such fragmentation by allowing only extents of the same size in the tablespace. That way, any freed extent is going to be exactly the right size for the next extent required in the tablespace. Avoiding different free space extent sizes requires different tablespaces To allow different-sized extents in your database without mixing them in the same storage space, you need different table-spaces. A variant of the fragmentation problem provides a fourth value for multiple tablespaces. Some segments are very unlikely to be dropped or truncated. For example, a mail-order business' CURRENT_ORDERS table unlikely will do anything but grow-or at least stay about the same size-as new orders are added and filled orders removed. On the other hand, you may have a data warehouse in which you keep the last five years' order records. If you build this table as a series of 60 month-long partitions, you'll be able to drop the oldest one each month as you add the latest month's records. Tables with different propensity to fragment free space The CURRENT_ORDERS table will never contribute to a fragmentation problem because its extents never go away. The data warehouse partitions, however, are dropped on a regular basis, so they'll have a high propensity to cause fragmentation. Thus, the fourth reason to keep segments in different tablespaces is to separate segments with a low, or zero, propensity to fragment space and those with a high likelihood of causing fragmentation. This way, the long-term objects, if they do need to grow, won't have to hunt around for free space of the right size. The fifth reason to use different tablespaces is to help distribute the data file reads and writes across multiple disk drives. You can use lots of different data files in a tablespace and place them on different disk drives, but you may not be able to control which extents are placed in which file after doing that. If you're lucky, the amount of disk access will be even across all the drives. If you aren't so lucky, you might have a situation where the two busiest extents in your database are in the same file. For example, if the mail-order house is going into its busy holiday season sale period, it will probably need to use the empty blocks at the end of the CURRENT_ORDER table for the additional orders. If the extent holding these blocks is in the same data file as the index blocks where the newest order numbers are being saved, you'll have excessive disk contention; each new order will use a block from the table extent and a block from the index extent. If you keep segments that will likely cause concurrent disk access (such as tables and the indexes on those tables) in different tablespaces, you can guarantee that the files making up the different tablespaces are stored on separate disk drives. A database-management issue is the final reason to use different tablespaces. During its life, a database will need to be backed up and possibly repaired if a disk crashes or otherwise corrupts data.

http://www.informit.com/content/0789716534/element_005.shtml (3 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

Tablespace damage can be pervasive If any part of a tablespace is damaged, the entire tablespace becomes unusable until the damage is fixed. If every segment belonging to an application were stored in a single tablespace and that tablespace was damaged, nobody could use that application. However, if you split segments that belong to different functional areas of the application (such as order entry and accounts receivable) and use different tablespaces for these, a data file problem may not be so intrusive. A failure with a data file in the accounts-receivable tablespace could be undergoing repair without any impact being felt by the order takers using the order-entry tablespace. Similarly, a tablespace containing tables that rarely change-such as lookup tables for state codes, part number and part name references, and so on-may not need regular backing up, whereas CURRENT_ORDERS may need very frequent backups to reduce the tablespace recovery time if there were a failure. Backing up tablespaces on different schedules To minimize the time it takes to back up less-often used segments, back up different table-spaces on different schedules. In fact, you can define a truly read-only tablespace to the database and then back it up only once when you've finished loading the required data into it. Of course, if you mix the static tables and the busy tables in the same tablespace, you have to back them all up equally often. SEE ALSO To learn more about objects, LOBs, and BFILEs,

Suggested Tablespaces
I recommend that every DBA create certain tablespaces for a production database. The reasons for these different tablespaces stem from the previous discussion. Let's begin with the SYSTEM tablespace, the only mandatory tablespace in every Oracle database. SYSTEM Tablespace Every Oracle database must have one tablespace-SYSTEM. This is where the user SYS stores the data dictionary information needed to manage the database. You should create additional tablespaces based on the expected use of your database. If you don't do this and use only the SYSTEM tablespace, you'll violate most of the reasons for using different tablespaces recommended in the previous section. Several other things will happen as well: Maintain the integrity of the SYSTEM tablespace You should never need to create any object directly in the SYSTEM tablespace, regardless of which userid you use to connect to the database. This table-space should be reserved for the recursive SQL executed behind the scenes as part of database creation or the execution of standard SQL statements.
q q

q q q q

You won't keep the segments owned by SYS out of harm's way by other users. You'll have to allow everyone who needs to create objects to take space from SYS in the SYSTEM tablespace. You'll cause fragmentation because all extents of all sizes will share the same space. You'll have a mix of segments with high and low propensity to fragment space in the same space. You can't easily avoid having high-usage extents stored on the same disk drive. You have a single point of failure for the whole database. If the data file in the SYSTEM tablespace is lost or damaged, the entire database will shut down.

I hope this list of possible problems has convinced you to use additional tablespaces, such as those described in the next few pages. Rollback Segment Tablespaces The tablespace for rollback segments will contain all the database's rollback segments with one exception-the SYSTEM rollback segment. It is maintained automatically in the SYSTEM tablespace. Keep your rollback segments separate from other database objects for a number of reasons: q They can shrink automatically and therefore create fragmented free space. Rollback segment shrinkage

http://www.informit.com/content/0789716534/element_005.shtml (4 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

Although you can define rollback segments to shrink by themselves if they grow too large, you can also use a special ALTER ROLLBACK SEGMENT command option to shrink them manually. SEE ALSO For details of the ALTER ROLLBACK SEGMENT command, They don't need quotas, so their space can't be managed by schema quotas. They're needed concurrently with data blocks during transaction processing, so they can lead to disk contention.

q q

Rollback segments can be defined to shrink themselves automatically if they grow larger than needed. Thus, they have a very high propensity for fragmenting the free space. Fortunately, a rollback segment is required to use the same size extents as it grows and shrinks, so it can reclaim any empty space it leaves behind when shrinking. Maintaining multiple rollback segment tablespaces If all the rollback segments in a single tablespace are sized identically, they can even claim the space released by the other segments. You shouldn't have many problems with space overuse or waste in a rollback segment tablespace with this arrangement, as long as they don't try to grow too much all at the same time. Should you need roll-back segments of different sizes, therefore, you should consider building a new rollback segment tablespace and keeping the larger ones in one tablespace and the smaller ones in the other. Another reason for using a tablespace for rollback segments is that you create them for use by anyone. Users don't need to have a space quota on the tablespace where the rollback segments reside. This allows the rollback segments to exist without having to contend for space with the segments that belong to an application (suddenly having their free space taken up by a table created with excessive storage requirements, for example). The final reason for keeping rollback segments in their own tablespace(s) is that you can help avoid contention for the disk. When a table is being modified, not only does Oracle modify the contents of the table block(s), but a set of rollback entries is also stored in a rollback segment. If the table and the rollback segment belonged to the same tablespace, the blocks being changed in each of them could be in the same data file on the same disk. Temporary Tablespaces Temporary tablespaces are for the temporary segments built by Oracle during sorting and related operations, when the amount of memory available to the server is insufficient to complete the task. The important characteristic of temporary segments is that Oracle creates and manages them without any input from the users. Consequently, unlike the other segments (except the bootstrap segment), there's no CREATE syntax to identify which tablespace the segment is created in, nor what size or number of extents to use. Neither is there any DROP command, or any other command, that lets you release the space from temporary segments. Default behavior of temporary segments Temporary segments obtain their storage characteristics solely from the tablespace definition. Hence, unless you have a tablespace for other types of segments that need identical storage characteristics to your temporary segments, you'll need a dedicated, temporary segment tablespace. Even if you have a tablespace that appears to be able to share storage characteristics for regular segments and for temporary segments, you may want to create a separate one for your temporary segments. The reason for this is that temporary segments are so named because of their default behavior. Because they're being used just to store intermediate values of an ongoing sort, they aren't really needed by the server process when the sort is complete. Hence, they can easily fragment the space in a tablespace as they are created and dropped. By default, temporary segments are dropped as soon as their work is complete. Obviously, the dropping of these segments on a transaction boundary makes them very prone to fragmenting free space. You can help mitigate the fragmentation problems by ensuring that the extents used by the temporary segments are all exactly the same size. The ephemeral nature of temporary segments by itself makes them candidates for their own tablespace. However, there's one further consideration-you can create a tablespace or alter an existing tablespace to contain only temporary segments. By doing this, you change the default behavior of temporary segments. Rather than drop the segment when it's no longer needed, the server simply records in a special data dictionary table that the extents in the segment are now free. Any other server needing to use temporary space can query this table and locate available temporary extents. In this way, the database will save the space allocated to temporary
http://www.informit.com/content/0789716534/element_005.shtml (5 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

segments indefinitely and assign it for use as needed by various servers. Reduce database overhead with TEMPORARY-type tablespaces The characteristic of preserving temporary segments in TEMPORARY-type tablespaces saves a lot of recursive SQL associated with creating and managing the space for temporary processing. In the long run, reducing recursive SQL speeds the processing of the entire data-base. It's essential that you have a tablespace dedicated to temporary segments if you want to take advantage of this behavior. You aren't allowed to place anything other than temporary segments in a table-space defined as a temporary type. As with the other tablespaces being discussed, you aren't limited to just one temporary tablespace (other than SYSTEM). You can add as many as you need to avoid contention for the space in a single temporary tablespace. Temporary space is assigned to users as part of the user definition. If you need to, you can subdivide your user community to access different temporary tablespaces. A final benefit of keeping your temporary tablespaces separate from other tablespaces is that because the use of temporary space is the result of processing being performed behind the scenes for the user, you don't need to assign space in temporary tablespaces to your users. You can keep your temporary tablespace(s) clear of other user-created segments by disallowing any storage on them. As with the SYSTEM tablespace, this will avoid problems that could occur should the required space be taken up by segments that don't belong in this reserved space. User Data Tablespaces For most of you, the largest amount of storage in your database will be taken by rows in your applications' tables. You should realize that these data segments don't belong in the SYSTEM tablespace, nor should they be stored with your rollback segments or your temporary segments. I recommend that you build one or more tablespaces to store your database tables. You would need more than one user data tablespace for a number of reasons, all related to the discussion in the section "Identifying Tablespace Uses": q Segment and extent sizes. It's improbable that all application tables will need to be the same size, or even have same-sized extents. To avoid fragmentation, you should place tables into tablespaces only with other tables that use the same extent size. q To allow you to manage them differently. If your database supports more than one application, you may want to protect them from one another by using separate tablespaces for them. In this way only one set of your users would be affected by a disk problem in a tablespace. Users of the applications supported in the unaffected tablespaces can continue their work while the repair work is done to the damaged tablespace. q To keep volatile tables away from static (or almost static) tables. This way you can organize your backup strategy around the needs of each tablespace-backing up the tablespaces with busy tables more frequently than those with tables less busy. For those tables that never change, or change very infrequently, you can make the tablespaces that hold them READ_ONLY. Such a tablespace needs to be backed only once following its conversion to this status. Managing tablespace extent size You may want to standardize your tables to three or four extent sizes. This will reduce the number of different tablespaces you'll need to manage while allowing you to realize the benefits of having all the extents in a tablespace be the same size. In particular, you won't have to concern yourself with the frequency at which extents are added and dropped. Such activity won't lead to poor space allocation because every dropped extent leaves free space exactly the same size a new extent would require.
q

To place your very busy tables in tablespaces different from each other. This way you can avoid the disk-drive contention that could occur if they share the same disks.

By the time you finish planning your user data tablespaces, you may have divided them for a combination of the reasons discussed here. It wouldn't be unreasonable to have two or three tablespaces holding tables with same-sized extents (each with a different backup frequency requirement), and another two or three with the same extent sizes containing tables that have the same backup requirements but have a high contention potential. Index Tablespaces Indexes on tables are often used by many concurrent users who are also accessing the tables as part of the same transaction. If you place the indexes in the same tablespace as the tables they support, you're likely to cause disk
http://www.informit.com/content/0789716534/element_005.shtml (6 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

contention; this would occur as queries retrieved index blocks to find where the required rows are stored and then retrieved the required data blocks. To avoid such contention, you should create a separate set of tablespaces for your indexes. As with tables, you may find that you need to size your index extents differently, and that you may have indexes supporting tables from different applications. Just as with your user data tablespaces, therefore, you should plan on building multiple index tablespaces to support different extent sizes and backup requirements and to maintain application independence in case of disk failure. If you use the Oracle8 partitioning option, you may need to revise your index and user-data tablespace design. In some cases it's beneficial to build locally partitioned indexes in the same tablespaces as the parent partition. This helps maintain the availability of the partitions during various tablespace maintenance activities. SEE ALSO For additional details on rollback segment management,

Understanding File Types: File Systems Versus Raw Devices

Some operating systems allow you to create raw partitions and use these for your Oracle database files instead of standard file-system files. In certain cases, raw devices can offer you improved performance. In addition, if you're using a UNIX system that doesn't support a write-through cache, you'll have to use raw devices to ensure that every write performed during a checkpoint or a log buffer flush is actually written to the physical disk. Raw devices and Oracle Parallel Server If you're going to use the Oracle Parallel Server option on UNIX or Windows NT, you must create all your database files-including the redo log files, control files, and data files-on raw devices. The various instances can't share the files if you don't do this. There are some drawbacks to raw devices: q On UNIX, sequential data reads from a raw device can't take advantage of read-ahead techniques. This means that a full table scan may be slower if other users are accessing different Oracle blocks on the same disk as the table being scanned. q Raw partitions aren't as flexible as file-system files. You're generally restricted to a certain number and, hence, fixed partition sizes on each disk. Also, you may not have permission as a DBA to create new raw partitions. This means that you have to work with a system administrator when you need to add a file. Although this may be easy enough when you're planning for a new database or a known expansion, it may not be convenient when you need to add or replace a data file in an emergency. Your system administrator may have to prebuild spare raw partitions for you to use in such emergencies. This approach, in turn, requires you to standardize on partition (and, therefore, on data file) sizes. As with extent sizes, you may want to simplify this approach by standardizing just three or four partition sizes across the whole database.

Understanding the Benefits of Striping Data

In some situations, no matter how well you've segregated your data into separate tablespaces, particularly busy tables or indexes will cause "hot spots" on one or more disks. To resolve this, you may need to build your data files by using an operating system striping mechanism. Such mechanisms include logical volume manager (LVM) software or redundant arrays of inexpensive disks (RAID). By striping the data file across many disks, you reduce the likelihood that any particular table or index will have all its busiest blocks on one single disk. If you decide to stripe your data files, you need to determine an appropriate stripe size. Optimal stripe sizes depend on a number of factors, the three key factors being the implementation of the striping software, the Oracle block size, and the type of database processing against the data. Stripe sizes for query-intensive data If your database is query-intensive- that is, the users are usually executing queries and very infrequently performing INSERTs, UPDATEs, or DELETEs-you need a stripe size that will support sequential processing of data. This will help queries that need to access indexes via range scans (many records with the same value or a set of values between an upper and lower bound) or tables via full table scans. You should set the stripe size for such databases to be a minimum of two times the value of the parameter DB_FILE_MULTIBLOCK _READ_COUNT and, if larger, to an integer multiple of this parameter's value. This will help ensure that all the blocks requested in a single read, when performing a full table scan, will be on the same disk, requiring only one disk read/write head to be moved to the required starting position.
http://www.informit.com/content/0789716534/element_005.shtml (7 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

Operating system or disk-striping mechanisms differ widely between vendors. Some offer large, guaranteed cache areas used for reads, writes, or both. A cache can overcome some performance slowdowns that can occur when you need to sequentially read blocks that are scattered across many different disks. Others, such as RAID, provide striping as a side benefit of various levels of disk failure resilience, but can slow certain activities to provide you protection against disk failure. Certain levels of RAID are better left unused for particular file types, such as those with large amounts of data written to them sequentially. For example, redo logs, while not part of a tablespace, may suffer a performance penalty if stored on RAID Level 5. If you have any tables that collect data in a similar sequential fashion, however, you should also try to avoid placing their data files on RAID Level 5 devices. Oracle block size is an important factor when striping data files that contain tables or indexes that will typically be accessed via random reads or random writes. You'll usually see this type of activity when the database is used primarily for transaction processing. In these cases, you should plan to make your stripe size at least twice the size of an Oracle block but, all things being equal, not too much larger. Whatever stripe size you choose, however, make sure it's an integer multiple of your Oracle block size.

Adding New Tablespaces

You automatically create your SYSTEM tablespace when you build your database. As discussed earlier in this chapter, you should add further tablespaces to meet your database's specific requirements. The following sections go over the process of creating tablespaces with various characteristics.

Creating a Tablespace
The very first tablespace you create is the SYSTEM tablespace, always part of an initial database creation. Additional tablespace creation isn't that much different from the SYSTEM tablespace creation. As with the CREATE DATABASE command, the CREATE TABLESPACE command uses a DATAFILE clause to identify the data file(s) and size(s) you want to associate with the tablespace. The syntax for the CREATE TABLESPACE command is as follows:

Identifies file(s) to be used and their characteristics Sets minimum size of used and free extents in table-space Determines whether certain SQL commands will avoid creating standard redo log entries Controls extent behavior for segments created without defined storage options Determines status of table-space after creation Defines tablespace to hold regular segments Defines tablespace to hold only temporary segments We'll examine the DEFAULT STORAGE clause in the following section. In the meantime, look at the DATAFILE clause in more detail. This clause can be applied to any tablespace's data files (including the SYSTEM tablespace), although most DBAs are content to use it simply to name and size the data file(s) for this tablespace. The DATAFILE clause's full syntax is as follows:

http://www.informit.com/content/0789716534/element_005.shtml (8 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

DATAFILE filename [SIZE integer [K|M]] [REUSE] [AUTOEXTEND OFF|ON [NEXT integer [K|M]] [MAXSIZE [UNLIMITED|integer [K|M]]]] Recall from the earlier section, "Understanding File Types: File Systems Versus Raw Devices," that you can use native file system files or raw partitions for your tablespace's files. The data file's name will therefore be a file-system filename, a raw partition name, or possibly a link name pointing to one or the other types of file. In the case of a file-system file, the file will be created for you unless you use the REUSE clause. In this case, Oracle will create the file if it doesn't already exist, but will overwrite an existing file as long as the SIZE clause, if included, matches the size of the existing file. If you don't specify REUSE and the file already exists, you get an error message, and the tablespace won't be created. The REUSE option can be destructive Be careful with REUSE; any current entries in the file will be overwritten and lost when it's implemented. A raw partition can always be reused, destroying its existing content-even if you don't include the REUSE keyword. If you name a raw partition (directly or via a link), the partition must already exist; otherwise, Oracle will attempt to create a standard file with the name of the partition. Because Oracle expects raw partitions to exist before being named in the CREATE TABLESPACE command, the REUSE clause really has no effect. The SIZE clause with raw partitions must be a few blocks smaller than the actual partition size; this allows space for operating system header information. Two operating system blocks are usually sufficient. Simplify your raw partition sizing You may want to keep your arithmetic simple when sizing raw partitions for Oracle files by allowing 1MB for the overhead in each partition. Thus, you would create a 101MB partition to hold a 100MB file. The AUTOEXTEND option determines whether a data file can grow automatically should a new extent be required by a segment and there's an insufficient number of contiguous free blocks. You don't have to use this clause (when you create the tablespace) to be able to grow your tablespace, as discussed later in the "Adding and Resizing Data Files" section. If you decide you want your files to be able to grow automatically, you should be aware of the impact of the following behaviors: q If there are multiple files with the AUTOEXTEND option in a single tablespace, the file chosen to grow when more space is needed will depend on a couple of characteristics. Oracle will try to extend the file that can be extended least to obtain the required space. If this results in a tie between two or more files, the one furthest from its maximum size will be extended. If this also results in a tie, the files will be extended in a round-robin fashion as more space is needed. q If the NEXT option isn't chosen, the files are extended one Oracle block at a time. q If you don't set an upper limit with the MAXSIZE option or use MAXSIZE UNLIMITED, the file can grow indefinitely until it reaches the limits of the physical storage device. q You can allow files in raw partitions to grow, but the data will overwrite the adjacent partition(s) if the partition size isn't large enough to contain the extended file; this destroys the contents and integrity of some other file. SEE ALSO For more information on creating a database, To learn about temporary segments and how they're used in TEMPORARY-type tablespaces,

Setting Default Storage Values

As noted in the earlier sections "Identifying Tablespace Uses" and "Suggested Tablespaces," a tablespace should ideally contain only segments with equal-sized extents. One way to help you maintain such a scheme is to assign to each tablespace the desired storage options as defaults. The CREATE TABLESPACE command's DEFAULT STORAGE clause is the means to achieve this. Your default storage values can be overridden! After a user has the privileges necessary to build segments, you can't prevent them from overriding the tablespace default storage values. You need to take on the responsibility of building all segments yourself if you don't want to risk this. To reduce the burden of work for your-self you can have the users build script files containing the required CREATE commands, which you can simply execute on their behalf. Of course, you should check that
http://www.informit.com/content/0789716534/element_005.shtml (9 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

Whenever a database segment such as a table, a rollback segment, or an index is created, a set of storage-related information is activated or stored with the segment definition. This storage information defines the size of the first extent belonging to the segment, the size of the second extent, the size of the subsequent extents, and the initial and the maximum number of extents that will be assigned to the segment. In the case of rollback segments, there's also a value associated with the optimal size of the rollback segment, which, if used, will cause extents to be dropped automatically if the overall size exceeds the desired maximum. SEE ALSO To learn more about managing rollback segments and the OPTIMAL storage option, Although each user who creates a segment can assign these storage values individually, they also can be inherited from the tablespace's definition. Temporary segments are a little different in that users don't get to create these; they're built as needed by the system on behalf of a user and, as such, always inherit the tablespace storage values. If you've built your tablespaces such that each one is designed to hold only one extent size, you can define the tablespace to provide this size by default. You can then advise those users who create segments (if it's someone other than yourself) that they shouldn't include the STORAGE clause in their CREATE statements. This not only simplifies their work, but keeps your tablespace extents defined as you planned. You define the inheritable storage values for a tablespace with the DEFAULT STORAGE clause. Here is the syntax for that clause: DEFAULT STORAGE (

Sets size of initial extent in bytes, with optional K or M to specify kilobytes or megabytes Sets size of second extent in bytes, with optional K or M Defines increase, measured as a percentage, by which each extent beyond the second will grow Sets number of extents each segment will be assigned when created Sets greatest number of extents that segment will be assigned You need to set INITIAL equal to NEXT and PCTINCREASE equal to 0 in order for the tablespace to create every extent, by default, with the same size. Remember that even though you set these defaults, every CREATE statement that builds a segment in the tablespace can override them. This is true even if you allow users to include a STORAGE clause simply to change the number of preliminary or maximum extents (MINEXTENTS and MAXEXTENTS). As soon as they can use a CREATE command, you can't restrict what's included in the related STORAGE clause. The following listing shows a command being used to create a tablespace with three data files, one of which is auto-extendible, and with a default storage clause to build all extents with 10MB of storage: CREATE TABLESPACE extra_room DATAFILE '/d1/oracle/exrm01.dbf' SIZE 1000M, '/d2/oracle/exrm02.dbf' SIZE 1000M, '/d3/oracle/exrm03.dbf' SIZE 1000M AUTOEXTEND ON NEXT 10M MAXSIZE 2000M DEFAULT STORAGE ( INITIAL 10M NEXT 10M PCTINCREASE 0) /

http://www.informit.com/content/0789716534/element_005.shtml (10 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

Tablespace Management
After you create your tablespaces, you may find that they aren't quite what you needed. To rectify this situation, you can drop and recreate the tablespace. In some cases you can modify it. The latter tends to be the easier solution if segments are already created in the tablespace, because dropping such a tablespace generally requires you to find a way to save and reload these segments.

Changing the Characteristics of a Tablespace

You can see a tablespace's current characteristics by viewing the data dictionary table DBA_TABLESPACES. Many of these characteristics can be changed with the ALTER TABLESPACE command. The syntax for the command is as follows: ALTER TABLESPACE tablespace_name option; Table 5.1 summarizes the options available with the ALTER TABLESPACE command. (You can only use one option at a time with the command.) Because some of the options are a little more complex than you might infer from Table 5.1, the following sections explain why you might want to use them. Table 5.1 ALTER TABLESPACE options Option: OFFLINE ONLINE BEGIN BACKUP END BACKUP LOGGING or NOLOGGING RENAME DATAFILE COALESCE MINIMUM EXTENT READ ONLY READ WRITE TEMPORARY PERMANENT DEFAULT ADD DATAFILE Purpose: Makes a tablespace unavailable for use and prevents access to its contents Returns a tablespace from OFFLINE to ONLINE accessible status Readies the files in the tablespace for hot backup Returns the tablespace's files to normal status following a hot backup Sets the default logging behavior of new objects created in the tablespace Identifies the new name of a data file to a tablespace when the file itself has been changed in the operating system Coalesces contiguous areas of free space into a single free extent Sets the minimum size of any extent, used or free, in the tablespace Prevents further writes into the tablespace Allows writes into the tablespace after it's read-only Converts a tablespace to one that holds only temporary segments Converts a temporary tablespace to a permanent one Changes the default extent characteristics assigned to any new STORAGE segments built in the tablespace Creates one or more additional data files for the tablespace

Removing Access to a Tablespace You may need to prevent access to a tablespace for a number of reasons. For example, you may want to back it up without users being able to change its contents, or you may need to perform maintenance or recovery on one of its data files. You can take a tablespace offline to prevent further read and write access. The ALTER TABLESPACE OFFLINE command that you use to accomplish this has three options: NORMAL, TEMPORARY, and IMMEDIATE. When you take a tablespace offline with the NORMAL option, Oracle immediately prevents further retrieval from that tablespace. However, it will complete a checkpoint on its data files before shutting it down completely; any changed blocks belonging to the tablespace still in the database buffer cache will be copied back to disk. This results in an internally consistent tablespace, so it can be brought back online at any time without any further processing. Bringing a tablespace back online

http://www.informit.com/content/0789716534/element_005.shtml (11 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

The ALTER TABLESPACE ONLINE command will bring an offline tablespace back online, provided that it was successfully check-pointed when it went offline and that all its files are currently online. If one or more of these conditions isn't true, the data file(s) will need recovery before the tablespace can be brought back online. Tablespace and data file recovery are discussed in Chapter 14, "Performing Database Recovery." The TEMPORARY and IMMEDIATE options of the OFFLINE command don't necessarily complete checkpoints. This can result in a tablespace inconsistent with the rest of the database that therefore may need media recovery when it's brought back online. To guarantee that the redo information required for this recovery is available when needed, the database must be running in ARCHIVELOG mode. The difference between TEMPORARY and IMMEDIATE is that the former will attempt to complete checkpoints on all the files, ignoring any not available for writes, whereas IMMEDIATE won't even attempt to process any checkpoints. Hot Backups of a Tablespace A hot tablespace backup is one made while access to the tablespace's data files continues. Even though making a copy of all data files associated with a tablespace may take a number of minutes, Oracle will allow users to read blocks from those files and modify those blocks, as well as allow DBWR to write the changes back to disk. This can result in apparent anomalies in the backup set. A table with blocks in two different data files could have some blocks in each file modified by a single transaction. The backup copy of one file could contain blocks as they were before the change, whereas the backup of the second file could contain changed images of other blocks. Oracle can resolve such anomalies by applying redo records to the backup files if they're used to replace damaged online files. To do this, the file needs to record the earliest time at which a block may have been changed but not copied into the backup file. This information is automatically available in a file header block, but normally this information will change over time. To prevent such a change from occurring, so as to lock in the time at which the physical backup begins, Oracle needs to freeze the header block for the duration of the backup. As the DBA, you need to issue ALTER TABLESPACE...BEGIN BACKUP before starting the physical backup of files in the tablespace. This will accomplish the freeze of the header blocks in the data files belonging to the tablespace as discussed earlier. You need to unfreeze these blocks when the backup is completed. You can achieve this with the ALTER TABLESPACE...END BACKUP command. Although you can place a number of tablespaces in backup mode simultaneously, you should understand one other characteristic of a tablespace's backup mode. While in backup mode, Oracle has to create additional redo information to guarantee data consistency within blocks. Block consistency during hot backups During a hot backup of a data file, it's possible for the operating system to copy different parts of an Oracle block to the backup medium in two separate read/write operations. If DBWR happened to write a new image of a block to the data file between the two operations, the backup would contain a fuzzy block image-part of the block would represent integral data at one point in time, while the remainder of the block would contain data from a different time. To ensure that a complete valid block image can be restored when recovering from this backup, Oracle places a complete block image into the redo log before any changes can be made to a block from a tablespace in backup mode. When recovering from the log, this valid block image is first copied over the possibly inconsistent block from the backed-up data file, and then the changes recorded in the redo are applied as usual. The redo logs needed to bring the data back to a consistent state must be available in order for the backed-up files to be useful in a recovery effort. To ensure this, you have to be running your database in ARCHIVELOG mode, which guarantees that all redos written to the online redo logs are copied elsewhere before the entries are overwritten by later transactions. You'll receive an error message if you try to place a tablespace into backup mode and you aren't archiving your redo. Controlling Logging Behavior A number of SQL commands can execute in Oracle without generating redo logs. These commands work with an existing set of data and therefore can be re-executed if they fail against the same data source. For this reason, you wouldn't have to rely on the existence of redo entries if there were an instance failure part of the way through the execution of the command. In addition, the SQL*Loader utility can run without logging because-again-the data source will still be available if the instance should fail before the load completes. These commands can be executed without the need for redo log generation:

http://www.informit.com/content/0789716534/element_005.shtml (12 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8
q q q q q q q q

INSERT, where data is being selected from another source CREATE TABLE...AS SELECT CREATE INDEX ALTER INDEX...REBUILD ALTER INDEX...REBUILD PARTITION ALTER INDEX...SPLIT PARTITION ALTER TABLE...SPLIT PARTITION ALTER TABLE...MOVE PARTITION

You can set the whole tablespace to a non-logging mode if your tablespace is going to contain many segments that you'll typically want to manipulate with these commands (and not generate redo entries). You must do this before you build the segments, however, because a segment will acquire only the tablespace's logging mode at the time it's created. You set the default logging mode for the tablespace with the ALTER TABLESPACE command, using the LOGGING or NOLOGGING option. When set, each new segment you create can accept this default behavior, or you can override it with the appropriate logging clause in the CREATE command. Moving Data Files There are generally two reasons to move a data file: q You're restoring a backed-up copy of the file following a disk failure and need to place the file on a different device or in a different directory structure than the original. q You've determined from monitoring database read/write performance that you have contention on certain disks. To solve this, you may need to move one or more data files to different disks. After you move a data file, you need to let the database know that the file has moved. You do this with a RENAME option of either the ALTER DATABASE or the ALTER TABLESPACE command. Generally, you use the former when the database is in a NOMOUNT mode, and you are in the process of recovering from media failure; you use the latter when you've completed a planned file move. In the latter case, you need to take the tablespace offline before physically moving the file and issuing the ALTER TABLESPACE...RENAME 'new_filename' TO 'old_filename' command. You can rename more than one data file in a single statement as long as they all belong to the same tablespace. Use a comma-separated list of filenames on each side of the TO keyword, ensuring that there's a one-to-one match between the names. For example, the following command will move three files from the /d1 device to three different devices: Oracle won't perform operating system file commands It's important to remember that renaming a file is a two-step process. Oracle doesn't physically move or rename the file at the operating system level; you are responsible for making this change yourself before issuing the ALTER TABLESPACE ...RENAME command. ALTER TABLESPACE prod_tables RENAME '/d1/prod02.dbf', '/d1/prod03.dbf', '/d1/prod04.dbf' TO '/d2/prod02.dbf', '/d3/prod03.dbf', '/d4/prod04.dbf' Coalescing Free Space Manually When there are multiple adjacent extents of free space in a tablespace, it can take longer for a new extent that spans these free extents to be created. If you monitor DBA_FREE_SPACE and notice that such free extents exist, you can manually coalesce them into one large free extent. You can issue the ALTER TABLESPACE...COALESCE command to combine the contiguous free extents in the tablespace on demand. Automatic free-space coalescing If you don't coalesce contiguous free space extents yourself, it will automatically be done for you by the background process SMON. The ALTER TABLESPACE... COALESCE option is provided because SMON may not work soon enough to be useful.
http://www.informit.com/content/0789716534/element_005.shtml (13 of 20) [26.05.2000 16:47:03]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

Avoiding Free Space Fragmentation One way to avoid having free space extents of various sizes is to prevent anyone from creating segments in the tablespace without your supervision. You can then ensure that they use extents of the same size for every segment. If this isn't an option, you can help minimize the problem by setting a "model" size for extents in the tablespace. This model size represents the smallest extent allowed and also controls the size of larger extents by ensuring that they are all integer multiples of the model size. If you decide to set such a model value for your tablespace, use the ALTER TABLESPACE...MINIMUM EXTENT command, providing an integer for the size, in bytes, of the smallest allowable extent. When MINIMUM EXTENT is set, every new extent added to the tablespace will be exactly the requested size, rounded up to the next Oracle block, or an integer multiple of that number of blocks. This sizing will override the tablespace's default storage clause, if necessary, as well as the storage options of the segment itself. Even manual extent allocations using such commands as ALTER TABLE...ALLOCATE EXTENT (SIZE...) will be controlled by the value MINIMUM EXTENT. Managing Query-Only Tables To avoid having to make backups of data files that contain non-changing data, you can define a tablespace and, consequently, its data files as read-only. Similar to putting a tablespace into backup mode, this freezes the related data files' header blocks. However, because there can be no changes to them, Oracle knows that these data files are current copies, no matter how long ago they were modified as read-only. Consequently, you can take a backup of such files and restore them, following media failure, at any time in the future without them needing any recovery information from the redo log files. Backup guidelines for read-only tablespaces You should back up the files in the tablespace as soon as possible every time you make a tablespace read-only; an earlier backup will still need to have redo applied to ensure that all changes before the change in status have been applied. Following a change to read/write again, you can still restore from the backup taken while it was read-only, provided that you have the redo generated following its change back to a read/write status. If you later need to make changes to one or more tables in a read-only tablespace, you have to make the tablespace accessible for writes again. You use the commands ALTER TABLESPACE...READ ONLY and ALTER TABLESPACE...READ WRITE to make these changes. Storage for Temporary Segments Temporary segments-used to complete sorts too large for the memory allocated to them-are ephemeral objects. They're created when needed and dropped when their work is done. In some databases-particularly query-intensive ones-the overhead of creating and dropping temporary segments can cause a significant performance problem. You can alter this default behavior by defining the tablespace where the temporary segments are stored to contain only this type of segment. Now, rather than drop a temporary segment when its work is finished, Oracle will preserve it for use by another sort in the future. If you didn't create the tablespace with this characteristic, you can issue the ALTER TABLESPACE...TEMPORARY to convert it to contain non-disappearing temporary segments. If the tablespace should happen to contain another type of segment, such as a table or index, you can't make this change. You can convert the tablespace if you need to add non-temporary segments or change the storage characteristics of the temporary segments in TEMPORARY tablespace. In this case you use the keyword PERMANENT in the ALTER TABLESPACE command. Any existing temporary segments will be dropped as they are when following their default behavior, and you'll be able to add any other type of required segment to the tablespace. You'll have to drop any of these segments ahead of time to reconvert the tablespace to TEMPORARY. Closing your database releases all temporary segments Temporary segment space isn't held over database shutdowns and restarts. Even the temporary segments stored in TEMPORARY-type tablespaces will have disappeared when you reopen a closed database. Modifying Default Storage Values The command ALTER TABLESPACE DEFAULT STORAGE allows you to change the default values assigned to the storage characteristics of segments created in the tablespace without their own, overriding, STORAGE
http://www.informit.com/content/0789716534/element_005.shtml (14 of 20) [26.05.2000 16:47:04]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

clauses. You need to take care when issuing this command for a couple of reasons: q This command affects only segments created after this change is made; it doesn't affect any existing segments. Specifically, if you have a table created with a default MAXEXTENTS value of 20, that table can contain only 20 extents-even if you change the tablespace default MAXEXTENTS to 40, 50, or even UNLIMITED. To change the storage characteristics of any existing segments, you have to alter each one individually; whenever a segment is created, the storage parameters for that object are stored in the data dictionary as part of that object's definition. Changing a tablespace's DEFAULT STORAGE changes only the tablespace definition, not the definitions of the objects within it. q If you've defined your tablespaces to contain extents of the same size, changing any of the INITIAL, NEXT, or PCTINCREASE default values causes any new object to build extents of sizes different from those of any existing segments. Therefore, unless you're prepared to deal with the possible fragmentation caused by different-sized extents, you shouldn't modify these particular values in anything other than an empty tablespace. Adding and Resizing Data Files You'll occasionally have to add space to an existing tablespace. This may be a planned or an unplanned occurrence: q Planned expansion is usually the result of an anticipated database growth over time in a system where the full complement of disk drives to support the growth wasn't available at database-creation time. It can also be the result of adding to an application more functionality that requires more rows or columns to be added to a table. q Unplanned expansion occurs when a segment, such as a table or an index, grows much larger than was anticipated in the database-design phase. This may be due to poor analysis or to a sudden change in the environment, such as an unanticipated doubling of orders for a specific product. For planned expansion, particularly those involving the addition of new disks, adding more data files is the best method for adding space to a tablespace. This allows you to add exactly the amount of space you need and to place it onto different disks from the existing files, thus avoiding possible disk contention. File-system files and raw partitions can be added by using the ALTER TABLESPACE...ADD DATAFILE command. As with the CREATE TABLESPACE command, you can add one or many files with the same statement. The file name(s) and size(s) specifications are just the same as in the CREATE TABLESPACE command discussed earlier in this chapter. You can also use additional data files, as just discussed, for an unplanned expansion. In such cases, you may not be able to place the files on new, unused disk drives; you may have to find whatever space is available in the disk farm for the time being. Also, if you need to use raw partitions, you'll have to be able to create them yourself or have the system administrator build them for you-unless you already have spares available. An alternative for an unplanned expansion is to let the data files grow themselves. This has to be done when you first add them to the tablespace, using the AUTOEXTEND clause with the CREATE or ALTER TABLESPACE commands' file specification. If you didn't set this option when you added the files to the tablespace, you can still increase the file's size by extending it manually. This command is ALTER DATABASE DATAFILE...RESIZE. (Notice that this is ALTER DATABASE, not ALTER TABLESPACE.) The RESIZE clause takes a single argument, indicating the number of bytes that you want the file to contain following successful execution of the command. This can either be a simple integer or an integer followed by K or M for kilobytes or megabytes, respectively. Shrinking oversized data files You can use ALTER DATABASE DATAFILE ...RESIZE to shrink, as well as to increase, the size of a data file. You can't reduce a file, however, unless there's empty space at the end of the file sufficient to remove the number of bytes needed to reach your desired size. The RESIZE option can't remove empty space from the middle of a file, and it won't remove blocks now assigned to a database object. The ALTER DATABASE DATAFILE...RESIZE command will manipulate only the space requested. It won't cause the file to expand, or shrink, automatically in the future.

Dropping Tablespaces
Although not a common requirement, you may need to drop a tablespace. There are a few reasons you might need to do this: q You no longer need the segments it contains for any further processing.

http://www.informit.com/content/0789716534/element_005.shtml (15 of 20) [26.05.2000 16:47:04]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8
q

There's enough corruption or damage to the contents of the tablespace that you want to rebuild it from scratch. You've moved its contents to another tablespace.

In order to drop a tablespace, it must not contain any rollback segments being used by an active transaction. If it contains any segments at all, you must use the INCLUDING CONTENTS option to force Oracle to drop these segments along with the tablespace. Dropping online tablespaces isn't recommended Although you can drop a table-space while it's still online, I advise you to take it offline first. This will avoid interference with ongoing transactions that are using the contents of the table-space and save you from drop-ping a segment that's really still being used. This is the DROP TABLESPACE command's full syntax: DROP TABLESPACE tablespace_name [INCLUDING CONTENTS] [CASCADE CONTRAINTS] You'll need the CASCADE CONSTRAINTS option if the tablespace contains tables being dropped with the INCLUDING CONTENTS option and these tables are the parents, via referential integrity constraints, of tables in another tablespace. SEE ALSO For a complete discussion of database ARCHIVELOG modes, To learn more about temporary segments, To learn about referential integrity constraints and the concepts of parent/child tables,

Extent Allocation
After you build your tablespaces, you or your users will use them to store various types of segments. Some of these will almost certainly be added by you and some will be automatically created by the Oracle kernel. The others may be created by you or by the users, but their maintenance and space management may still be under your control in either case. Part of the work involved in managing segment space allocation should be completed during the physical design of your database because it's related to the number and arrangement of your tablespaces. This topic is discussed earlier, in the section "Space Management Fundamentals"; the discussion that follows here assumes that you've already decided what type of segment is being placed where and concentrates on how new extents are added to these segments after they're created, or how unneeded space can be retrieved from segments to which it has already been allocated.

Comparing Dynamic and Manual Extent Allocation

Every type of Oracle segment (except the bootstrap segment, which is fixed in size at database-creation time) can grow automatically by default. The following sections examine how each segment type can have its growth controlled and how best to manage any required growth. Temporary Segments We begin this discussion with temporary segments because in many ways these are the easiest to manage-you have so little control over them. Temporary segments are created as needed by Oracle while it's processing SQL statements on behalf of a user's process. Temporary space is generally required by a process if performing a sort operation, although some types of joins and related activities also use temporary segments. This temporary disk space is used only when the SQL statement has insufficient memory in which to complete its processing. All extent allocation to temporary segments is dynamic. In other words, it occurs automatically, without any specific commands from users. This is true whether a new temporary segment is being created or new extents are being added because the original size was insufficient to complete the task. Due to this completely automatic behavior, users in no way can give explicit instructions about the sizing or the number of extents in a temporary segment. Oracle always uses the storage information defined in the DEFAULT STORAGE clause of the tablespace for its temporary segments. In Chapter 6 "Managing Redo Logs, Rollback Segments, and Temporary Segments," you can find guidelines on how to determine good storage values for temporary segments.

http://www.informit.com/content/0789716534/element_005.shtml (16 of 20) [26.05.2000 16:47:04]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

By default, as soon as the operation that required the disk space is finished, the temporary segment is dropped and the blocks used by its extents are returned to tablespaces as free blocks. It's this behavior that gave temporary segments their name; they only acquire space for a short time and then return it. Another option you should consider for your temporary tablespaces is creating or converting them to the TEMPORARY type. This will prevent Oracle from dropping temporary segments in the tablespace following the completion of the related SQL statements. Instead, the extents used in the segment are tracked in the data dictionary and made available to any server process that needs temporary space. New extents are added to the segments only if all the current extents are now in use by one or more users. Closing your database releases all temporary segments Temporary segment space isn't held over database shutdowns and restarts. Even the temporary segments stored in TEMPORARY-type tablespaces will have disappeared when you reopen a closed database. By not forcing users to create and recreate temporary segments each time they're needed, their work can be completed much faster. In fact, you can save your users a lot of time when you first create a TEMPORARY-type tablespace by prebuilding all the extents the tablespace can hold. You can do this by performing a massive sort (if you have a table or set of tables large enough to join), or by running large sorts in a number of concurrent sessions. Make sure that the userid you use for these sorts is allotted to the temporary tablespace you're planning to populate. Another benefit to using TEMPORARY-type tablespaces for your temporary segments is that Oracle enforces the use of same-size extents. All extents in such tablespaces are built based on the value of the NEXT parameter in the DEFAULT STORAGE clause. Rollback Segments As the DBA, you should create and manage rollback segments. You initially create a rollback segment with two or more extents and with extent sizes taken from the tablespace default values or from CREATE TABLESPACE's STORAGE clause. The behavior of the extents allocated to rollback segments is of interest here. SEE ALSO You can find detailed information about creating rollback segments on Rollback segments store information that would be needed if a transaction were to roll back. Every part of a single transaction must be stored in the same rollback segment, and many transactions can share the same segment. In most cases, transactions generate the same amount of rollback information, so when a rollback segment reaches a certain size, its space is sufficient to support all the needed concurrent transactions. As these transactions complete, the space they were using is recycled and made available to new transactions. However, if the database gets very busy or suddenly needs to support one or more very long-running transactions, a rollback segment may need to grow by adding one or more extents. As with temporary segments, this allocation is dynamic; users have no control over it. Rollback segments have one unique characteristic of space management not possessed by any other type of segment: They can shrink in size by dropping unnecessary extents. Suppose a rollback segment grew by adding extents in response to an unusual combination of concurrent long-running transactions. If before this it could handle its work load without additional space, it should be able to do so again without the need for additional space. If it's sharing a tablespace with other rollback segments, this space might be better used by one of the others, maybe also for a sudden increase in work. You can cause a rollback segment to return to this preferred size whenever it exceeds it by setting an OPTIMAL parameter value. OPTIMAL is a special parameter Whereas all the other storage parameters for a rollback segment can be inherited from the tablespace definition, OPTIMAL must be set with the STORAGE clause of the CREATE ROLLBACK SEGMENT or the ALTER ROLLBACK SEGMENT command. Data and Index Segments Segments designated to store table or index data can be created by you or by userids responsible for the applications that will use them. These segments can inherit all their tablespace's storage characteristics, just some of them, or none of them. When created, their storage characteristics can be changed for the most part; only the INITIAL and MINEXTENTS values are fixed for the life of the segment. If a data or index segment runs out of space in its current extents, one of two things can occur: A new extent

http://www.informit.com/content/0789716534/element_005.shtml (17 of 20) [26.05.2000 16:47:04]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

will be added by Oracle dynamically, or the SQL statement that required the extra space will fail. There are a number of reasons dynamic allocation could fail: q There may be insufficient space in the tablespace, and none of the data files can autoextend. q There may be space in the tablespace, but there isn't a sufficiently large extent of free space to hold the required extent. q The segment may already contain the MAXEXTENTS number of extents. Adding space to allow a new extent to be created automatically when the failed SQL statement is re-executed was discussed earlier in "Adding and Resizing Data Files." This would address the two first causes of dynamic space allocation failure. Another option for handling the second problem is to change the value of the segment's NEXT storage option, causing it to create an extent that fits into a remaining free extent. A third option would be to allocate the extent manually. You use the ALTER TABLE ALLOCATE EXTENT clause to do this. The complete syntax is as follows:

Identifies which freelist group will manage blocks in extent (used for databases that use Oracle Parallel Server option) Sets extent size, regardless of table's storage values Identifies into which data file extent will be placed This command has one additional benefit over changing the NEXT value. If you want, you can execute it a number of times, each time choosing a different size for the extent and a different data file into which it goes. This will allow you to prebuild extents that precisely fit the available free space until you have sufficient space allocated; this allows work on the segment to continue while a more permanent solution, such as additional disk space, is found. You can take advantage of manual extent allocation with the ALLOCATE EXTENT option for reasons other than overcoming space limitations. For example, you may want to build a segment in a tablespace with many data files so that you guarantee that blocks from each data file will be used by the segment. To do this, you can create the table or index with a single extent and then use the DBA_EXTENTS data dictionary view to find out which data file contains this extent. Then, by successive use of the ALTER TABLE...ALLOCATE EXTENT command, you can place an additional extent into each data file belonging to the tablespace. A final note on manual extent allocation If you don't provide a size when manually allocating an extent, the extent will be sized as though it were created dynamically. If you use the SIZE clause, however, it won't override the dynamic sizing that would have occurred. If a table were going to build its next dynamic extent with 1,000 blocks and you manually add an extent of just 50 blocks, the next dynamically allocated extent would still acquire 1,000 blocks.

Releasing Unused Space

Occasionally you'll build a segment far larger than you need it to be-perhaps because the initial estimates made during the analysis and design phase were wrong, or because the nature of the application changed. If you need to regain the unused space, you have a variety of options. First, when dealing with the case of rollback segments, you can use the ALTER ROLLBACK SEGMENT command to change the value of OPTIMAL. As long as you don't try to shrink it to a size less than its original size (that is, size of extent 1 + size of extent 2 + ... + size of extent MINEXTENTS), the rollback segment will, as it's used in future transactions, return to this size. As long as it doesn't consistently run out of space, it will attempt to maintain this size even if it temporarily grows beyond it. For a one-time fix, you do have the option of issuing the following command: SHRINK can cause free extents of unequal size
http://www.informit.com/content/0789716534/element_005.shtml (18 of 20) [26.05.2000 16:47:04]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

This command will remove partial extents. Therefore, even if you've carefully built your tablespaces and segments to have equal-sized extents, the end result of this command can be an extent of a size smaller than planned and a piece of free space larger than the expected extent size. ALTER ROLLBACK SEGMENT...SHRINK [TO integer [K|M]] For table, cluster, and index segments, you can remove unused space with the ALTER command's DEALLOCATE UNUSED clause. This command removes any unused extents and blocks as long as the original extents, set with the MINEXTENTS value in the CREATE command, aren't involved. You can even remove some empty space with the optional KEEP clause. This will save some allocated space to allow for some future growth without further extent allocation. A second option to remove excessive space from a table-one that will preserve extent sizes-is to move the data into a temporary table, drop all the extents (other than the original ones) from the table, and then move the rows back into it. The following code shows a session that performs exactly these actions on the UNFILLED_ORDERS table. The key commands are the CREATE TABLE...AS SELECT and TRUNCATE commands. CREATE TABLE temp AS SELECT * FROM unfilled_orders / TRUNCATE TABLE unfilled_orders / INSERT INTO unfilled_orders SELECT * FROM temp / DROP TABLE temp / To drop unused space from an index, you can simply use the ALTER INDEX command's REBUILD option. The only restriction you need to be concerned with when using this command is that the original and the replacement index copies will temporarily have to exist at the same time. This means that you'll need space for the new version of the index to be built in the target tablespace, which may not be the same as the current one.

Defragmenting Free Space

If you have a database that for whatever reason doesn't follow the guidelines discussed in this chapter about maintaining equal-sized extents in a tablespace, you may find yourself with a badly fragmented tablespace. This tablespace contains a lot of free space, but each piece, or extent, of free space is only a few blocks big, too small to be usefully added to any of the segments in the tablespace. For tablespaces of type TEMPORARY, this will only occur should the default storage settings have been changed before the tablespace is completely filled. To fix it, you can alter the tablespace to be a PERMANENT tablespace again. This will cause all the temporary segments to be freed over time and new ones built in their place. When all the odd-sized extents have been removed, you can convert the tablespace back to type TEMPORARY and allow the new segments to grow back to the necessary sizes by using fixed-sized extents. For rollback segment tablespaces, your best option is to temporarily provide additional rollback segments to give you a chance to drop and recreate the current rollback segments, choosing appropriate storage values in your CREATE command. If you already have rollback segments available in another tablespace, you may be able to make the changes without adding temporary ones. If you plan to do this, try to drop and recreate the problem rollback segments during a period of low use. That way users aren't held up because an insufficient number of rollback segments are available to support them. The most difficult fragmentation problems to fix are those associated with tables and indexes. If you can afford to drop all the indexes in the tablespace temporarily and then rebuild them, this is the easiest way to solve the problem. However, if the tablespace contains tables or other types of segments besides the indexes, you need to deal with the larger problem. Similarly, if the indexes can't all be dropped, you don't have a simple method to solve the issue. Need to defragment a tablespace?

http://www.informit.com/content/0789716534/element_005.shtml (19 of 20) [26.05.2000 16:47:04]

informit.com -- Your Brain is Hungry. InformIT - Managing Your Database Space From: Using Oracle8

If you have to defragment a table-space, I strongly recommend that you reconsider your tablespace usage before reloading anything. You may want to change all the segment storage clauses so that they all have equal-sized extents, or you may want to add more tablespaces to meet the design suggestions offered earlier in this chapter. As your database grows in size, the inconvenience to you and to your users will increase should you need to perform future defragmentation processing. Fragmented tablespaces containing tables, or tables and other types of objects, are very difficult to handle. Some third-party tools are available. Without them, you're going to use a tool to store the contents of the tablespace in some type of temporary storage, drop the tablespace contents, and then restore the original contents by using new segments with appropriate sizes. Oracle offers the Export and Import utilities to help you do this. You can also build your own tools to unload data, table definitions, and the like, and then use a combination of SQL, SQL script files, and SQL*Loader to reload the tablespace. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_005.shtml (20 of 20) [26.05.2000 16:47:04]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Managing Redo Logs, Rollback Segments, and Temporary Segments

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q

Database Management Structures Managing Redo Log Files

r r r r r

Sizing Your Redo Logs Determining the Number of Redo Log Groups Determining the Number of Redo Log Members Adding Redo to Your Database Dropping Redo Logs and Handling Problem Logs Determining the Number of Rollback Segments Sizing Your Rollback Segments Adding Rollback Segments Creating a Rollback Segment PUBLIC Versus PRIVATE Rollback Segments Altering Rollback Segments Dropping and Shrinking Rollback Segments Sizing Your Temporary Tablespaces Setting Storage Options for Your Temporary Tablespaces Managing Your Temporary Segments

Managing Rollback Segments

r r r r r r r

Working with Temporary Segments

r r r

Database Management Structures

To most database users, tables are the important elements, with indexes being a second component they might consider as useful. In an Oracle environment, however, some structures are essential to the efficiency and integrity of the database. You need to know how to build and manage these components for your database to run smoothly and to allow your users to be able to work with their tables in an orderly manner. Oracle8 can be forgiving, but don't count on it

http://www.informit.com/content/0789716534/element_006.shtml (1 of 18) [26.05.2000 16:47:20]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

Although you can run your database without paying attention to your redo logs and temporary segments, you'll almost certainly pay a performance penalty by ignoring them. In most cases, these penalties will be severe and may cause your data-base to fail to meet your users' requirements. Rollback segments are a little more intrusive, at least if you pay any attention to Chapter 5's recommendations to use multiple tablespaces in your database, because you'll find that the default database structures won't support DBA or user-created segments in new tablespaces. The three structures you look at in this chapter are redo log files, rollback segments, and temporary segments. If you've already read Chapter 5 "Managing Your Database Space," you'll already be aware of some characteristics of these last two. You have been exposed to the concept of redo log files if you've read Chapter 2 "Creating a Database." Although the purpose of these structures will be touched on in this chapter, the emphasis in the following sections is to help you design, build, and manage them to the benefit of your database.

Managing Redo Log Files

One of the slowest operations performed in a relational database is the process of transferring data between disk storage (where it's maintained for long-term safety) and memory (where it needs to reside to be accessed and modified by database users). Although providing you with a number of parameters to tune memory and help avoid unnecessary disk reads, Oracle has its own mechanism-known as deferred writes-to reduce the overhead of unnecessary disk writes. Basically, it means that Oracle blocks that have been modified by user processes while sitting in memory will be written back to disk when they can be written most efficiently. This approach takes into account how recently the last change was made, how many blocks can be written with a single write operation, and how much of the space occupied by these blocks is needed for further disk reads. One key issue to understand about deferred writes is that blocks aren't moved back to disk just because the changes on them have been committed by a user, and that blocks containing uncommitted changes are just as likely to be written to disk as blocks with committed changes. In other words, the act of issuing a COMMIT doesn't override the deferred write considerations. This, in turn, leads to a situation where neither the contents of memory nor the contents of the data files represent a coherent picture of the database transactions. With such a design, it's essential that Oracle provide a mechanism to restore the database to a consistent state should the contents of memory be lost. Of course, disk loss also needs to be protected against, but you can use a number of techniques, discussed in the chapters in Part V, "Backing Up Your Oracle Database," to help ensure that such a loss can be recovered. There's no method to "back up" the contents of memory in case it should suffer a failure. The importance of a COMMIT When users issue the COMMIT command and receive an acknowledgment that the command is processed, the changes to the transaction are considered "permanent." This doesn't mean that they can't be changed in the future, but that the assigned values should be available to all other users of the database until they're changed again. This requires that the DBMS guarantees that committed data won't be lost by any fault of the database itself (database vendors can't control the quality of the media on which their databases are stored). Oracle has to employ a method-redo logs-to ensure that the contents of unwritten blocks containing committed data can be restored should unexpected memory loss occur. Understand that Oracle can lose its memory area for many reasons, not just the loss or corruption of memory due to a chip failure, but also due to an unexpected operating system failure, loss of a key Oracle process that prevents further processing (and, hence, further access to memory), or a sudden shutdown of the Oracle instance without normal shutdown activity (a SHUTDOWN ABORT). Anytime Oracle stops operating without being able to complete a requested shutdown command, it's said to have suffered "instance failure." In any such case, there's no opportunity to copy block images from memory down to disk nor to reload blocks that contain uncommitted changes. Instance failure will require that, when the instance is restarted, an instance recovery be performed to reapply all changes associated with committed transactions and to remove all uncommitted changes. Implied SHUTDOWN ABORT The server manager command STARTUP FORCE is used to cycle an instance-that is, it stops the current instance and starts a new one. However, it performs an implicit SHUTDOWN ABORT command to stop the running instance. Consequently, this command will cause an apparent instance failure, requiring an instance recovery as part of the requested startup process. You shouldn't use this command to restart an instance quickly, but only when you would normally need to use a SHUTDOWN ABORT.

http://www.informit.com/content/0789716534/element_006.shtml (2 of 18) [26.05.2000 16:47:20]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

The redo log files contain the information needed to complete an instance recovery. They will allow the recovery operation to re-execute every command that produced part, or all, of a committed database change, whether the affected database blocks were copied back to disk before the memory failure or not. Similarly, they contain enough information to roll back any block changes that were written to disk but not committed before the memory loss. Without delving into the details of what really goes into a redo log and how the recovery process works-information that's explained fully in the Oracle8 Server Administrator's Guide and the Oracle8 Server Concepts manuals provided as part of the Oracle documentation-the following sections explain what you need to consider when building your database and preparing it for your users and applications.

Sizing Your Redo Logs

If you make a mistake in the arithmetic in your check register, you can go back to the last time you balanced your account with a bank statement and redo the steps to get a valid current balance. But if you don't know when you last had a valid balance, you can't. Similarly, for Oracle to apply redo information following an instance failure, it must be able to find a starting point at which the status of all data and transactions was known before the failure. This starting point is known as a checkpoint. Very simply, whenever a checkpoint occurs, Oracle forces every changed block now in memory to be written to disk to ensure that the redo records associated with those changes are stored in the redo log. When these steps are complete, a special checkpoint marker record is placed into the redo log. If the instance fails before the next checkpoint completes, the recovery operation knows that the log file and the data files were synchronized at the previous checkpoint. Any changes made to a data block since that checkpoint was completed are recorded in the redo log entries written after its checkpoint marker. Therefore, recovery can begin at the most recent checkpoint marker record. With this checkpoint mechanism in place, entries older than the most recent checkpoint aren't needed. To conserve disk space, Oracle will reuse the redo logs. When all the logs are filled up, the LGWR process simply starts to write over the first redo log file again. Figure 6.1 shows how this circular use of the redo logs occurs. For this to be successful, Oracle has to ensure that a checkpoint has been taken before reusing a redo, and this is achieved as part of the processing that occurs at a log switch. At this time, LGWR will automatically begin writing new redo information into the next available log file and will also initiate a checkpoint. Therefore, your redo log size has a direct bearing on the frequency of these default checkpoints. Figure 6.1 : LGWR reuses the redo log files in a circular fashion, overwriting old records as it goes. You need to concern yourself with this because the amount of data written to a log file between checkpoints affects the length of the instance recovery process (the act of reapplying changes following an instance failure). If a log file holds a day's worth of transactions, it could easily take a day to reapply them after an instance failure. Assuming that it may take a few minutes to a few hours to research and fix whatever problem caused the instance failure, the total recovery time in such a case could be more than a day. Most businesses can't afford to have a key database unavailable for that long, unless it's a planned period of downtime during which other plans have been made to carry on with essential processing. Instance failure is, by definition, an unplanned event. Of course, if your database is being used primarily for queries (which, by their nature, don't change the contents of the blocks being read), your redo log could be quite small and still not fill up in a 24-hour period. Only when new data is added, or old data is changed or removed, will redo entries be written to the log, and this is very infrequent in read-intensive databases. Time to apply redo during instance recovery If you're running Oracle8 on a multi-CPU platform, it's likely that your users are creating transactions and the related redo log records in parallel. When your database needs instance recovery, only one process-the background process SMON-will be reading the redo logs and reapplying the changes made by your users. This serial processing can take substantially longer, in elapsed time, than the transactions took when being processed by multiple users concurrently. Another factor to consider is that you can set parameters to cause additional checkpoints to be performed automatically as a redo log fills up. The LOG_CHECKPOINT_INTERVAL parameter sets the number of blocks that will be written to a redo log file before a checkpoint is forced to occur, whether or not the file is full. If a redo log file is 1,000 blocks in size, setting LOG_CHECKPOINT_INTERVAL to 250 will cause a checkpoint when it's one-quarter, one-half, and three-quarters full. Although such a setting will reduce your instance recovery time for this redo log, it may be inappropriate should you decide to reduce your redo log size, for

http://www.informit.com/content/0789716534/element_006.shtml (3 of 18) [26.05.2000 16:47:20]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

example, to expedite archiving (as discussed a little later in this section), and then forget to reset this parameter. There are a number of considerations when setting checkpoint frequency, as the following two examples demonstrate: q A customer proposed using an Oracle database to track the airborne spread of environmental pollutants during emergencies. During crises, it would be imperative to ensure that the most current information was readily available to emergency response teams. The information was to be loaded from a number of tracking stations in real-time. In case of an instance failure, the customer wanted to restore operations as quickly as possible to continue to collect the most current data and to make it available to users in the field. To minimize downtime during instance recovery, it was recommended that checkpoints be taken every few seconds. Although not necessary for most users, a number of customers that need up-to-the-second real-time data have also instituted checkpoints many times a minute. q On the other extreme, some commercial customers who experience heavy volumes of transactions for a few hours a day, such as bank counter service operations, have elected to avoid checkpoints during these hours by building large redo logs that can handle a full day's business without switching-and hence without causing any checkpoints. They force a checkpoint every day immediately before the work period to ensure that they have a complete log file available for that day's load. Ideally, you don't want checkpoints to occur more than once every 15 to 20 minutes, and much less frequently if the database is mainly processing queries. The problem with starting checkpoints too frequently is that a number of very active blocks will still be in use between checkpoints. However, they will have to be written out at each checkpoint. Redundant writes waste the disk I/O bandwidth. You may want to experiment with log file sizes after you finish reading about other considerations, such as archive logging (discussed in a little bit), to come close to the ideal size. Parallel server processes might speed up instance recovery You can set the RECOVERY _PARALLELISM parameter in your initialization file to an integer value higher than 1 to allow SMON to enlist that number of parallel server processes for recovery. You must also start this number of processes by using the PARALLEL_MIN _SERVERS parameter. These processes will apply the redo in parallel during instance recovery. You may not see significant improvement in recovery time, however, because the parallel server processes must still apply the redo in sequential order, so they're likely to be contending for disk read access to the redo logs as well as contending for space in the database buffer cache. Before leaving checkpoints, you should be aware of one other factor: To perform instance recovery, a checkpoint marker must be available to indicate the start point. If you have two log files and, for whatever reason the checkpoint following a log switch doesn't complete until the second log fills up, the only checkpoint marker is in the first log file. If Oracle began to write redo records into this first log file again, there's no guarantee that this remaining checkpoint wouldn't be overwritten, leaving no starting point for an instance recovery. Consequently, Oracle will stop writing further redo entries until the checkpoint process completes and the new marker record can be written. If no redo log entries can be written, Oracle can't preserve the integrity of database changes because block images in memory can't be guaranteed to be recoverable. So, rather than let unrecoverable changes occur, Oracle stops any further transaction processing. To the users, the database will appear completely frozen. Of course, after the checkpoint completes, work will continue as normal. You may have to size your redo logs large enough to avoid this problem because a database that freezes out user activity isn't going to meet performance standards. A second mechanism that may affect your redo log file size decision is whether you're going to archive your log files. Although this is another topic that belongs in the backup and recovery discussions in Chapter 12, "Understanding Oracle8 Backup Options," we'll take a quick look at it here. Normally, when a redo log is filled and another one is being written to, the contents of the first log are of no use following the completion of the checkpoint started at the log switch. When the other log file fills up, Oracle can safely begin writing over the contents in the first one. Similarly, because a new checkpoint will be under way, the data in the second log file will soon become unnecessary for instance recovery, so Oracle can switch back to it when the current log fills. Now consider data file backups. They can't be made continuously, so the restoration of a backed-up data file will almost certainly cause old block images to be placed back into the database. Transactions completed since the backup was made won't be represented. However, if you could keep every redo entry made since the data file backup was made, restoring the blocks in that data file would require nothing different than restoring them following an instance failure.

http://www.informit.com/content/0789716534/element_006.shtml (4 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

Oracle offers the capability to "archive" your online redo log files so that they can be preserved for this very purpose. As each redo log fills up, Oracle still switches to the next one and starts a checkpoint, but also marks the completed redo log file for archiving. Either you or a special background process, ARCH, will copy the redo log to a special location where it can be saved for as long as needed. Users and log file switches when archiving Be sure to understand the pros and cons of archiving before deciding whether to use it, and then be prepared to monitor your system for problems with postponed log switches until archived copies are made. Not only does processing for current users come to a halt during such times, but also new users attempting to connect to the data-base will be prevented from doing so. This can make the problem very visible to your user community. Even if your users don't let you know, the alter log for your data-base will show you when you have log-switching problems due to tardy archiving. When you place your database into the mode that requires completed log files to be saved to an archive location, Oracle becomes very adamant that this work be performed. In fact, it won't let the redo activity switch back into a log file until that file has been safely archived. So, if your files are very big and take too long to archive (particularly if they're being copied to a slow disk drive) or so small that they fill up faster than they can be copied, you can run into problems. If the logs can't be switched because the archiving isn't done, no more log records can be written. Only when the archive is complete can work continue again. During the time that Oracle is waiting for the archive to finish, your users are experiencing the same situation when checkpoint completion was delayed. To them, the database is stuck and they can't get any work done. You may therefore have to adjust your log file size to ensure that the archiving process will complete sooner than the next switch.

Determining the Number of Redo Log Groups

Besides changing the size of your redo log files to avoid problems with checkpoints and archives not completing in time, you can add the amount of redo that's written before a log file is needed again by adding more log files. If it takes half an hour to fill each log file, and you have two such files, it will take an hour before you'll fill both logs and need to reuse the space in one of them (refer to Figure 6.1). If this isn't always enough time to complete your checkpoints and archives, you have a serious performance problem and should turn to Chapter 20 to find out about redo log tuning. However, until you can resolve the problem, you could increase the length of time before a log file is reused by adding one more file. Generally, databases with smaller log files and high DML activity will tend to have peak periods during which their log files fill faster than checkpoints or archives complete. By adding more log files, you increase the time taken to fill up the entire set of logs. This allows time for the checkpoint and archive work to catch up from the overloads during the peak periods. A suggested option for sizing online redo logs One approach to redo I've seen work successfully is to keep as much online redo log as can be filled between database back-ups. This way, should you have to restore a data file from a backup, the redo needed to recover it up to the point of failure will be in the online redo log. The recovery operations can access the online redo logs more directly than they can archived logs, so the recovery time will be reduced. You need to find a good balance between the size of your log files, which affects default checkpoint intervals (and hence instance recovery times), and the number of your log files, which with the size determines how long each checkpoint and archive has to complete. In general, you're better off having too much online (as opposed to archived, or offline) redo log rather than too little. Too little will likely cause at least an occasional pause while a checkpoint or an archive completes; too much will simply waste disk space. Oracle requires you to have a minimum of two redo log files. You can have up to 255, unless your operating system sets a lower maximum number. When you create a database, you can reduce the maximum number of redo log files you will be allowed to create by setting the optional parameter value, MAXLOGFILES, of the CREATE DATABASE command. You can even set an initialization parameter, LOG_FILES, to limit the number of log files that an instance can access.

Determining the Number of Redo Log Members

Each log file is given a log group number, either by you as you add them or automatically by Oracle. We refer
http://www.informit.com/content/0789716534/element_006.shtml (5 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

to log files with a group number because a group can contain more than one file. Each member of a group will be maintained by Oracle to ensure that it contains the same redo entries. This is done to avoid making a redo log a single point of failure. When your log groups contain only one member, you risk having the database become unusable if you lose a redo file. Recall from the earlier section, "Sizing Your Redo Logs," that at least one checkpoint completion marker must be available somewhere in your redo logs. If only one such marker happened to be in a set of logs and the file containing the marker was on a disk that crashed, you would no longer have a way of performing instance recovery. This jeopardizes your database and so Oracle, on detecting a missing log file, will stop processing any more transactions and perform a shutdown. Oracle mirrors versus operating-system mirrors for redo logs There has been much discussion within Oracle and with Oracle's business partners about the pros and cons of using Oracle's multiplexing versus using a mirrored disk controlled by the operating system. The biggest benefit to Oracle mirrors is that they work on any operating system and on any disks; the biggest disadvantage is that Oracle insists that each available copy is written to before it considers a flush of the redo buffer complete. This synchronous write process can be slower than operating-system mirrors. However, as disk subsystems become faster and add intelligent buffering capability, the latter difference becomes less of an issue. My best advice at this time is to use Oracle mirroring if you have no other option, and to experiment with Oracle and operating system-mirroring if you can. If each log file is paired with a copy of itself and that copy is on a different disk, a single disk failure won't reduce the database to an unusable state. Even if the only checkpoint record was in the file on a crashed disk, its copy would still contain a valid version of it. Oracle will know to avoid the bad disk for future writes and for any further archiving activity. The term Oracle uses for copied sets of redo logs is multiplexing. You're strongly encouraged, therefore, to multiplex every log group with at least two members. Depending on the criticality of your systems, you may want even more. Rarely do you need to go beyond three members per group; in fact, with more than that, you're likely to experience performance problems due to the time it takes to write out of the copies of each redo block. If you can mirror your log files at the operating-system level, you can also use mirroring to guard against a single disk loss. If you rely on operating-system mirroring alone, you still run the risk of having Oracle shut itself down if you lose a disk. System mirrors aren't visible to Oracle, so it may think it has lost its only copy of a log file if the primary disk crashes. System mirroring is a good way to create three- or four-way mirroring, however. Create each Oracle log group with two members, and then mirror either one or both members.

Adding Redo to Your Database

When you created your database, you created at least two redo log groups, the minimum number required to start an Oracle database. You may have created more than that, and you could have created each group with one or more members. This section looks at the commands used to add more log groups or more log members to an existing group. You'll see that the syntax is very similar to the log definition portion of the CREATE DATABASE command. One option you can use when creating a redo log group allows you to identify a thread number. The thread number is useful only in a parallel server (multi-instance) database, so we won't examine its usage here. For further details on this option, see the Oracle8 SQL Reference Manual and the Oracle8 Parallel Server Administration manual. Syntax conventions used in this book Throughout this book, square brackets in command syntax indicate optional clauses and an ellipsis ([...]) indicates a clause that can repeat. Another convention used is the | character, which indicates that you choose between one item or the other, not both (for example, choose either K or M). When you actually use commands, don't include the brackets, ellipses, or | character. The general syntax for creating a log group with a single member is as follows: ALTER DATABASE [database_name] ADD LOGFILE [GROUP [group_number]] filename [SIZE size_integer [K|M]] [REUSE] This code is for a multimember group:

http://www.informit.com/content/0789716534/element_006.shtml (6 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

ALTER DATABASE [database_name] ADD LOGFILE [GROUP [group_number]] (filename, filename [,...]) [SIZE size_integer [K|M]] [REUSE] The database name is optional if it's included in the parameter file (as the DB_NAME parameter) for the instance. Otherwise, you need to identify the name with which the database was created and which is stored in the control file. If you omit the group clause (the keyword GROUP and the group number), Oracle will assign the next available group number for you. Every group must have a unique number to identify it. The filename can be a file system filename (which should be fully qualified with a path name), a raw partition name, or a link. In the multimember case, you should put the filenames inside a pair of parentheses and separate the names with commas. You must include a SIZE or a REUSE clause. You can include both for file system files, as long as any existing file is the same size as the specification. For file system files, you must provide a size if the file doesn't already exist, and you must include the REUSE keyword if the file does exist; the command will fail if either condition is violated. For raw partitions, the REUSE keyword is meaningless because the new contents will always be written over the contents of the partition; therefore, it makes no difference whether you include it. You must include the file size, however, to avoid using the whole partition-two blocks of space must be reserved in each partition for operating-system information-or possibly writing beyond the partition boundaries. The K and M represent kilobytes and megabytes, respectively. Without either, the size_integer represents bytes. The SIZE and REUSE options in database redo log groups If you're creating a log group with multiple members, include the SIZE or REUSE keyword only once for all members of the group. They must all be the same size because they'll all contain the same data. This means-unless you're using raw devices-that if one file exists, they must all exist so that the REUSE option is valid for each named file. If some exist and some don't, you'll have to create the group with only those that exist (or only those that don't) and add the others as additional members. I show you how to do this a little later. No matter how you create them, all the files in a redo log group will have to be same size. Listing 6.1 shows a script file with three commands, each creating a new redo log group. Listing 6.1 Create new redo log groups 01: 02: 03: 04: 05: 06: 07: 08: 09: ALTER DATABASE ADD LOGFILE D:\ORANT\DATABASE\log10.ora SIZE 100K / ALTER DATABASE ADD LOGFILE GROUP 6 (E:\DATABASE\log6a.ora, F:\DATABASE\log6b.ora) SIZE 10M / ALTER DATABASE ADD LOGFILE GROUP 5 (E:\DATABASE\log5a.log, F:\DATABASE\log5b.log) REUSE /

Numbering of code lines Line numberings were included in Listing 6.1 and other code listings to make discussion about this code easier to reference. The numbers should not be included with any command-line commands, as part of any Oracle scripts, or within SQL statements. On line 1 of Listing 6.1, the first redo log group will be created with a single member in the group, and the group's number will be assigned by Oracle. Group 6 will have two members, and the group is assigned its group number in the command on line 4. In these first two commands, Oracle will create all new files. Redo log group 5, as created by the command on line 7, will contain two members, both of which will replace existing files. Adding one or more new members to an existing group can be done by identifying the group number (the simplest syntax) or by identifying the group with a list containing the full path names of all the current members. The syntax for the former when adding just one more member is ALTER DATABASE database_name ADD LOGFILE MEMBER

http://www.informit.com/content/0789716534/element_006.shtml (7 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

filename [REUSE] TO GROUP group_number

Different numbers of members per log group Oracle doesn't require that you use the same number of log file members in each group. In fact, because you can add a new member or members to only one group at a time with the ALTER DATABASE command, you couldn't start mirroring your log files by adding a new member to each group unless they could exist with different numbers of members, at least temporarily. However, even though you could run your database with two members in one redo log, three in another, just one in a third, and so on, I don't recommend this practice. After you decide how many mirrored copies make sense for your requirements, you should use that number in all groups. This way, you won't experience periods of different performance or have to worry, should you lose a disk drive, whether you've lost a single-copy redo log or just one of a mirrored set. The database name is optional, as when adding a new group. The group number must refer to an existing group. The filename must be a fully qualified file system name, a raw partition, or a link. The REUSE keyword is needed only if you're using a file system file that already exists, in which case it must be the same size as other files in the group. A SIZE clause isn't needed because every member of the group must be the same size as the existing member(s). The syntax for using the existing filename(s) to add a single member is as follows: ALTER DATABASE database_name ADD LOGFILE MEMBER filename [REUSE] TO [filename] | [(filename, filename, (,...)] Everything is as described earlier except that for a group with a single member, the filename alone is used in place of the GROUP clause, whereas a comma-separated list of the existing member's filenames (enclosed in parentheses) is required if the group already has more than one member. In either case, the filenames must be fully specified. To add multiple members to a group withi in the same command, you simply change the filename clause to read as follows in either version of the statement: (filename, filename, (,...)) [REUSE] The use of REUSE is, as before, required if the files already exist.

Dropping Redo Logs and Handling Problem Logs

The most likely reason you would want to drop a redo log file is because you want to replace it with one of a different size. Once in a while you may drop a log file because you've determined you have more online redo than you need. This is rarely beneficial, however, unless it's the only log file on the disk, because you generally shouldn't share disks where you're writing online redo logs with other file types. You may also need to drop a redo log member when you're experiencing performance problems due to too many multiplexed copies, or because you want to replace one or more members with operating-system mirrored copies. To drop an entire log group, the following must be true: q At least two other log groups will be available after the group is dropped. q The group isn't currently in need of archiving. q The group isn't the current redo group (the one to which log entries are now being written). If these conditions are met, you can drop the entire log group with an ALTER DATABASE command that identifies the group. As with the ADD LOGFILE MEMBER option of this command (discussed in the preceding section), you identify the group with its group number, its member's filename, or with a list of the filenames of its current members: ALTER DATABASE database_name DROP LOGFILE GROUP group_number | filename |

http://www.informit.com/content/0789716534/element_006.shtml (8 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

(filename, filename (,...)) The database name is needed only if the parameter file used to start the instance doesn't include the DB_NAME parameter and the ellipsis ([...]) shows a repeatable field. You can drop one or more members from an existing log group with the DROP LOGFILE MEMBER variant of this command. You can't drop all members with this command, however; you must use the preceding command to drop the group as a whole. The syntax for dropping a group member is ALTER DATABASE database_name DROP LOGFILE MEMBER filename where the database name has the same requirements as previously discussed, and the filename must be fully qualified, as with all files discussed in these sections. Once in a while, a redo log group may become damaged to the point where the database can't continue to function and you need to replace the redo group with a clean file or set of members. If the damaged log group isn't yet archived or the log group is one of only two log groups in the database, however, you aren't allowed to drop it. Creating a third log might not help because Oracle will continue to attempt to use the damaged log before moving on to the new one. In such cases, you need to simulate dropping and recreating the log with the CLEAR LOGFILE option of the ALTER DATABASE command. After you do this, you may need to perform a brand new backup of your database because there may be a break in the continuity of your archived logs, and you may have removed the only checkpoint record in the online redo. If you do have to perform an emergency replacement of an online redo log, use the following command: ALTER DATABASE database_name CLEAR [UNARCHIVED] LOGFILE group_identifier [UNRECOVERABLE DATAFILE] where database_name and group_identifier follow the same characteristics as described earlier for the DROP LOGFILE option. The UNARCHIVED clause is needed if the group was awaiting archiving before being cleared, and the UNRECOVERABLE DATAFILE option is required if the log would have been needed to recover an offline data file. To find out about the current status of your redo logs, you can query various dynamic performance views. The V$LOGFILE view will show the names of the members of each redo log group and their status. In this view, NULL is a normal status, INVALID indicates that the file is unavailable, DELETED shows that the file has been dropped, and STALE is used when a file is a new member of a group or doesn't contain a complete set of records for some reason. The V$LOG and V$THREAD provide more detailed status information and include records of the archive and system change numbers related to the redo files. Also, the view V$LOG_HISTORY is used mainly by parallel server databases for recovery operations. SEE ALSO How to set up redo log archiving for your database, Learn about tuning your redo logs for checkpoint and archive processing, More about the alert log and the types of messages it can provide, such as log file switches delayed by checkpoints or archiving,

Managing Rollback Segments

Rollback segments perform two basic functions, both somewhat related: q Allowing changes to be rolled back (as the name suggests). This activity restores block images to the state they were in before a change was made. The change could be a row INSERT, UPDATE, or DELETE, but could also be simply a change in the header portion of the block to reflect a change in the status of a transaction. q Providing read-consistent images of blocks. These are copies of blocks restored to look just as they did when a query or series of queries began. Known as "undo blocks," they allow a query to read blocks being changed by other users, preventing it from seeing uncommitted changes (dirty reads) or changes that occur while the query is executing (inconsistent reads). To perform its function, a rollback segment stores a before image of a column, row, or other block element before the change is applied to the block. By using the address of this changed data, also stored with the before image, a rollback or read-consistency operation can overlay the changed information with this record of what it

http://www.informit.com/content/0789716534/element_006.shtml (9 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

looked like before the change. Dirty reads A "dirty read" is a query that returns value from a row that's part of an as-yet uncommitted transaction. If the transaction is subsequently rolled back, the query has returned a value that's never really been stored in the database. An inconsistent read occurs when a query reads some blocks before a transaction changes them and other blocks after the same transaction changes those. During a rollback operation, the before image data is applied directly to the data block image where the transaction had made its changes. Rollbacks can occur for a number of reasons, including, but not limited to the following: q The user or application issuing a ROLLBACK command q A single statement failing after making some changes q A transaction failing because the user is unexpectedly disconnected from the database q An instance being recovered following an instance crash; the transactions incomplete at the time of the failure are rolled back as part of the instance recovery mechanism In some cases, particularly the latter, the blocks that need rollback information applied may be stored only on disk rather than in memory. When a read-consistent block image is needed, Oracle first copies the block into a different memory location inside the database buffer cache. The original block image can continue to be manipulated by any active transactions that need to modify it. Oracle then applies the rollback information to the copy of the block, called the "undo block." In some cases, a long-running query may encounter a block that has been changed by multiple transactions subsequent to the start of the query. In such a case, the undo block will be further modified by applying the before images from each transaction until the undo block resembles how the original block looked when the query began. The query will then read the undo block as opposed to the "real" block. Rather than allowing rollback segments to grow indefinitely, Oracle reuses the blocks that contain before images of completed transactions. Over time, the entire rollback segment is recycled many, many times as new transactions find space for their rollback entries. This reuse of space is controlled rather than haphazard, however. For the read-consistent feature to work, the before images needed by a query need to be available for the whole duration of the query. If a new transaction simply reused any available rollback block, it could be the one needed by an executing query. To help avoid this, the space is used in a circular fashion. The oldest before images are overwritten first. To simplify the code to support this activity, a couple of rules are applied to rollback segments: q Only one extent is considered to be the active extent. When a new transaction needs to store a before image, it's assigned to a block within the active extent. As soon as the active extent fills up, the next extent is made the active extent. Transactions that run out of space in their assigned block will be given a second block in the active extent or, if none are available, will be assigned a block in the next extent, making it the active extent. q When an extent fills up, if the next extent still contains at least one block with before images from a still-active transaction, that extent isn't used. Instead, Oracle builds a brand new extent and makes it the active extent. In this way, all the blocks in the extent with the active transaction are left available for queries that might need their contents to build undo blocks. This behavior is shown in Figure 6.2. Figure 6.2 : Oracle uses rollback segment extents in a circular fashion unless they're all busy, in which case it builds a new one. By cycling through the extents or building new ones when necessary, a block in, say, extent 1 won't be overwritten until all the blocks in all the other extents have been reused. This allows before images to remain available for the longest time possible, given the current size of the rollback segment. Preserving the before images for queries is important because, if a query needs a before image that's not available, the query can't continue. Without the before image, the query can't reconstruct the block in question to look as it did at the query start time and it terminates with an error message: ORA-1555 - Snapshot too old. The message ORA-1555 - Snapshot too old is usually a warning that at least one of your rollback segments is too small to hold enough records to provide read consistency. If it occurs very infrequently, however, it may simply indicate that a report, or other query-intensive program, ran into a busy period of transaction processing that it usually avoids. If rerunning the problem program succeeds, you may not need to change your rollback segment sizes for this infrequent occurrence.

http://www.informit.com/content/0789716534/element_006.shtml (10 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

The ORA-1555 error message One cause of the ORA-1555 problem needs to be solved by the application developer rather than by a change in the rollback segment sizing. The error occurs if a program is making changes to many rows in a table by using an explicit cursor data-either in PL/SQL or a 3GL language with Oracle precompiled code-to read through the data and additional cursors to make changes to the required rows. If these individual row changes are committed, the query cursor needs to build read-consistent images of the affected blocks. While this may not involve much rollback information itself, it does require the query to find the transaction entry information in the header blocks of the roll-back segments involved. It's the sheer number of transactions, not their size, that causes ORA-1555 errors in this type of program. The following sections discuss characteristics of transaction rollback and read consistency that you need to consider when determining the size and number of your database's rollback segments.

Determining the Number of Rollback Segments

A rollback segment uses the first block in its first extent to build a list of information about the transactions assigned to it. This list can contain only a fixed number of entries because the block size is itself fixed. Therefore, if your database needs to support lots of concurrent transactions, you should consider adding more rollback segments. Although Oracle will try to balance the workload between the rollback segments as new transactions begin, if every rollback segment is now supporting its maximum number of transactions, new transactions will be forced to wait until others complete. The SYSTEM rollback segment When a database is created, the default rollback segment is created in the SYSTEM tablespace, and it takes the default storage parameters of the tablespace. You cannot drop this rollback segment. It is used by Oracle for recursive SQL. For performance reasons, you really shouldn't allow each rollback segment to support its maximum number of transactions. If you do, you'll overwork the header blocks of the rollback segments. Not only do new transactions have to place their entries on these blocks, but the status of ongoing transactions has to be recorded, including information about the extents they're actively using. Finally, when a transaction completes, the status of the transaction-including a system change number for committed transactions-has to be recorded in the transaction's slot in the header block. There's no absolute rule as to the average number of transactions you should strive for per rollback segment. In most cases, the best number is between 4 and 10. For longer-running transactions that consume more rollback space, you should use a lower number in this range. Similarly, for shorter transactions, a higher number is a better target. You can get an idea about what your transaction mix looks like by examining the following dynamic performance tables. When you know the total number of concurrent transactions, divide this number by the appropriate value between 4 and 10, based on transaction size, to find a good number of rollback segments to try. You can always adjust this number later. SEE ALSO Determining if you have too many or too few rollback segments by examining contention statistics, Additional information on rollback segment performance,

Sizing Your Rollback Segments

After you determine how many rollback segments you need, you need to figure out how large they should be. For most databases, there is a perfect rollback segment size for the normal workload. At this size, rollback segments will contain sufficient space to hold the rollback data needed for all the concurrently active transactions at any point in time. However, they achieve this without being oversized, which would waste space. Once in a while you may have a special job or program that needs an extra-large rollback segment. If so, you can build one such rollback segment and leave it available at all times, or else leave it offline until needed. Assigning transactions to specific rollback segments

http://www.informit.com/content/0789716534/element_006.shtml (11 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

To ensure that a transaction uses a specific rollback segment, you can take other rollback segments offline, or you can explicitly assign the rollback segment with a SET TRANSACTION USE ROLLBACK SEGMENT rollback_segment_ name command. If you have concurrent transactions or the segment is needed by an application that runs outside your control, explicit assignment is better. The SET TRANSACTION command must be executed before every transaction if the same rollback segment is needed for each one. You generally don't have to worry about making your rollback segments too small because, like other segments, they can grow automatically as more space is needed. This growth does depend on whether the segment has reached the maximum number of extents you've defined for it and on the amount of room remaining in the tablespace where it's stored. See the following section, "Adding Rollback Segments," for details on how to set the extent maximums and tablespace allocation. I don't recommend letting Oracle take care of rollback segment growth for you for a couple of reasons: q Any such dynamic growth will slow down the process that incurs the overhead of finding the required free space and allocating it to the segment. q Studies performed at Oracle have shown that rollback segments perform best when they have between 10 and 20 extents. If you rely on Oracle to add extents as needed, you may have segments well outside these ideal limits. Another problem with automatic growth is that, once in a while, something will occur that will make it grow far larger than is typically necessary. One example I have encountered was a program that, following a minor change, got itself into a processing loop that caused it repeatedly to update the same few records without committing the changes. As a result, the rollback segment handling the transaction kept growing until its tablespace ran completely out of space. At that point, the transaction failed. When that happened, the space in the rollback segment taken up by the runaway transaction entries was freed up for use by subsequent transactions-but the rollback segment was now almost the size of the tablespace. When a different transaction, assigned to another rollback segment, needed more space for its entries, it failed because its rollback segment had insufficient room to grow. By using the OPTIMAL entry in the STORAGE clause of the CREATE or ALTER ROLLBACK SEGMENT command, you can hone in on the best size for your rollback segments. The OPTIMAL value will cause the rollback to perform a special check when it fills up its current active extent. If the sum of the sizes of the current extents is greater than OPTIMAL, rather than just look to see if the next extent is available to be the active extent, the server checks on the one after that, too. If this one is also available, the server will drop the next extent rather than make it current. Now, if the total rollback segment is at its optimal size, the current extent becomes the one following the dropped extent. But if the total size is still greater than OPTIMAL, the extent following this one is checked for availability and the same process is repeated. Eventually, the rollback segment will be reduced to optimal size by the deletion of extents, and the next remaining extent will become the current extent. Rollback segment extents are dropped in a specific order The extents are dropped in the same order that they would have been reused. This activity results in the oldest rollback entries being dropped, preserving the most recent ones for use by ongoing queries. You can query the dynamic performance table V$ROLLSTAT to determine how many times a rollback segment has grown through the addition of new extents (the EXTENDS column value), and how many times it has shrunk (the SHRINKS column value). If these numbers are low, or zero, the rollback segment is either sized correctly or it may still be larger than needed. You can adjust the value of OPTIMAL downward and check the statistics again later. If they're still low, or zero, your extent may still be oversized. However, if they have started increasing, it means the rollback segment needs to be larger. If the grow-and-shrink counts are high when you first look at the table, the rollback segment has always been too small. Problems to look for when decreasing rollback segment size

http://www.informit.com/content/0789716534/element_006.shtml (12 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

When reducing the size of a roll-back segment, you need to monitor your users' queries to ensure that the number of Snapshot too old messages doesn't increase. Remember that queries may need rollback information long after a transaction is finished. You can't just size your rollback segments to make them as small as the transaction load requires if this will interfere with standard query processing. Even a report program that runs once a month may require you to maintain larger roll-back segments than your records of rollback segment growth and shrinkage would indicate are needed. If the program fails only once or twice a year, the cost of rerunning it may not be as expensive as the cost of the extra disk space needed to support larger rollback segments. But if it fails almost every month, you may need to increase your roll-back segment sizes. Due to the automatic nature of the extent additions and deletions, you don't have to recreate a rollback segment that's the wrong size-you can control it with the OPTIMAL value once you find its stable size. As mentioned earlier, however, a rollback segment performs optimally when it has between 10 and 20 extents. This number provides the best balance between the need for transactions to find available space and the availability of required rollback entries for queries needing read-consistent data. Of course, based on the discussion of space management in Chapter 5 we're talking about rollback segments where all the extents are the same size. If your rollback segment's ideal size corresponds to this preferred number of extents, you can leave it as now defined. If the number of extents is below 10 or much above 20, however, you should consider dropping it and re-creating it with around 15 equal-sized extents, such that its total space remains the same.

Adding Rollback Segments

If you've just created a new database, there's only one rollback segment, SYSTEM, and it won't support transactions on segments outside the SYSTEM tablespace. From the discussions in Chapter 5 you should be building all your application tables and indexes in alternative tablespaces, and so you'll need additional rollback segments to support transactions against them. If you have a running database, based on statistical results from your monitoring and tuning activity, you may determine that you need to add one more rollback segment to the existing set. In either case, you follow exactly the same steps. Add rollback segments 1. Create the rollback segment. 2. Bring the rollback segment online. 3. Alter the parameter file to bring it online automatically whenever the instance starts.

Creating a Rollback Segment

Listing 6.2 shows the syntax of the command to create a rollback segment. Listing 6.2 Create a rollback segment 01: 02: 03: 04: 05: 06: 07: CREATE [PUBLIC] ROLLBACK SEGMENT segment_name [TABLESPACE tablespace_name] [STORAGE ( [INITIAL integer [K|M]] [NEXT integer [K|M]] [MINEXTENTS integer] [MAXEXTENTS integer] [OPTIMAL NULL|integer[K|M]] ) ]

On line 1, PUBLIC causes the rollback segment to be public rather than private. (This distinction is discussed in the next section, "PUBLIC versus PRIVATE Rollback Segments.") segment_name is a valid Oracle name. On line 3, INITIAL is the size of the first extent, in bytes (default) or in K kilobytes or M megabytes. NEXT on line 4 is the size of the second and subsequent extents, in bytes (default) or in K kilobytes or M megabytes. Line 5 shows MINEXTENTS, which is the number of extents (minimum two) included in the rollback segment at creation time and the number of extents that must always belong to the segment. MAXEXTENTS on line 6 is the largest number of extents the segment can acquire. Although MAXEXTENTS can
http://www.informit.com/content/0789716534/element_006.shtml (13 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

be set to the value UNLIMITED, this isn't recommended for rollback segments. If you've sized your rollback segments correctly, they shouldn't need to grow much larger than this; unlimited growth would result from erroneous processing. Such processing could fill up the available space, restricting the growth of other rollback segments performing valid work, and would take as long to roll back, when it finally ran out of space, as it did to build all the rollback entries in the first place. Until the rollback of this transaction is completed-which could conceivably take many, many hours, if not days-the space consumed by the rollback entries can't be freed. No PCTINCREASE option for rollback segments Every extent, other than the first, must be the same size. With a non-NULL value for OPTIMAL, any extent with no active transactions assigned to it can be dropped if, by so doing, the total segment size will still be greater than the OPTIMAL size. The initial extent is never dropped, however, because it maintains the transaction table in its header block. Also, only the extents that have been inactive the longest are dropped. If there are four inactive extents but an active one between the third and fourth of these, only the first three will be dropped. This is to avoid removing records that might be needed for read-consistent queries. On line 7 is OPTIMAL, which determines how the rollback segment can shrink. A value of NULL prevents the rollback segment from shrinking automatically; a size (in bytes, kilobytes, or megabytes) causes the rollback segment to shrink automatically by dropping inactive segments. OPTIMAL must be set to a value no smaller than the sum of bytes in the first MINEXTENTS. This can be computed from the formula OPTIMAL >= INITIAL + (NEXT * (MINEXTENTS - 1))

PUBLIC Versus PRIVATE Rollback Segments

The main reason Oracle supports public rollback segments is to help DBAs who manage multiple Oracle instances running against a database with Oracle Parallel Server. The management of rollback segments on some hardware platforms running Parallel Server is almost impossible without them being public, mainly because you don't have to name the rollback segment in a parameter file to make it come online when an instance starts up. Instead, each instance takes one or more segments from the pool of public rollback segments as its own. Although you can use public rollback segments in a non-parallel database, you're encouraged to use private rollback segments. By naming the rollback segments you want to make active in an instance in your parameter file, you have full control over which ones are active. While you can name public rollback segments in the parameter file, two other parameters-TRANSACTIONS and TRANSACTIONS_PER_ROLLBACK_SEGMENT-can also bring additional ones online, if available. By using private rollback segments, you're guaranteed that only those named in the parameter will be brought online automatically at instance startup. How Oracle activates public roll-back segments for an instance Oracle evaluates the quotient of TRANSACTIONS and TRANSACTIONS_PER_ ROLLBACK_SEGMENT when it starts up. If these parameters were set to 210 and 25, respectively, the result of this calculation would be 8.4. Oracle rounds up this value to the next integer and attempts to acquire this many rollback segments for the instance. First, it counts the number assigned by the ROLLBACK_SEGMENTS parameter and then calculates how many more are needed. To continue the example, sup-pose that five rollback segments were named in ROLLBACK_SEGMENTS; 9 minus 5-or 4 more-would be needed. If at least four public rollback segments aren't yet assigned to an instance, the current instance will take four of these and bring them online for itself. If there are fewer than four, it will bring as many online as are available. If no public rollback segments are available, the instance will continue to run with just the five named in the parameter file. You can take private and public rollback segments offline and return them to online status while the instance is running, as discussed in the following section.

Altering Rollback Segments

Before working with a rollback segment, you may need to determine its current status. This can be a little confusing due to the number of different data dictionary tables you might need to examine to get the full picture. Table 6.1 shows the various types of rollback segment status and characteristics, along with the data dictionary table and column that contain this information.

http://www.informit.com/content/0789716534/element_006.shtml (14 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

Table 6.1 Identifying the status of a rollback segment Status ONLINE OFFLINE PENDING OFFLINE DEFERRED PRIVATE PUBLIC Data Dictionary Table DBA_ROLLBACK_SEGS V$ROLLSTAT DBA_ROLLBACK_SEGS V$ROLLSTAT V$ROLLSTAT DBA_SEGMENTS DBA_ROLLBACK_SEGS DBA_ROLLBACK_SEGS Table Column STATUS STATUS STATUS STATUS STATUS SEGMENT_TYPE OWNER (= SYS) OWNER (= PUBLIC)

In an online state, the rollback segment is available for use and may have active transactions running against it. In an offline state, the rollback segment is idle and has no active transactions. A pending offline state is a transition state between being online and being offline. When you alter an online rollback segment to be offline, it won't accept any more transactions, but will continue to process any current transactions until they complete. Until these are all committed or rolled back, the rollback segment remains in the offline pending state. A deferred rollback segment holds rollback information for transactions that can't complete because the tablespace to which they need to write has gone offline. These transactions will have failed due to the loss of the tablespace, but they can't be rolled back because the blocks in the offline tablespace can't be read or written. To be able to complete the necessary rollbacks when the tablespace comes back online, the associated rollback entries are stored in the SYSTEM tablespace in deferred rollback segments. Although not truly a status, Table 6.1 also includes an entry for the PRIVATE and PUBLIC rollback segment descriptions so that you know how to identify which is which. As you can see, this is shown indirectly in the OWNER column of the DBA_ROLLBACK_SEGS table, where an entry of SYS indicates that it's a private rollback segment and an entry of PUBLIC shows it to be a public rollback segment. The ALTER ROLLBACK SEGMENT command lets you change the status of a rollback segment manually. The full syntax for this command is ALTER ROLLBACK SEGMENT segment_name [ONLINE|OFFLINE] [SHRINK [TO integer [K|M]]] [STORAGE (storage_clause)] The keywords ONLINE and OFFLINE simply take the rollback segment between the basic states. As discussed earlier, a rollback segment may not go completely offline immediately; it may have to wait until pending transactions complete. If you're taking a rollback segment offline in preparation for dropping it, you may need to wait until it's completely offline, as shown in V$ROLLSTAT. You're not allowed to take the SYSTEM rollback segment offline for any reason. The SHRINK keyword causes the rollback segment to shrink to its optimal size, or to the size provided when you execute the ALTER ROLLBACK SEGMENT command. As with automatic shrinkage, you can't reduce the size to less than the space taken by MINEXTENTS. The storage clause of the ALTER ROLLBACK SEGMENT command is identical to its counterpart in the CREATE ROLLBACK SEGMENT statement, with the following provisos: q You can't change the value of INITIAL or MINEXTENTS. q You can't set MAXEXTENTS to a value lower than the current number of extents. q You can't set MAXEXTENTS to UNLIMITED if any existing extent has fewer than four blocks (and you're advised not to use this value anyway, for the reasons discussed earlier).

Dropping and Shrinking Rollback Segments

At times, you may want to change the physical structure of a rollback segment, such as alter its basic extent size or move it to another tablespace. You may even want to remove a rollback segment because you no longer think you need it. If you have a rollback segment that has grown much larger than the optimal size and want it reduced to the optimal size as soon as possible, you can shrink it yourself. To remove a rollback segment, you must first take it offline, as discussed in the previous section. Recall that even if the command to take it offline works, you may not be able to remove it. Until all the transactions now assigned to the segment complete, the status won't be permanently altered to offline, so you won't be able to drop it. Of course, after you change its status, no further transactions will be assigned to the rollback segment.

http://www.informit.com/content/0789716534/element_006.shtml (15 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

As soon as a rollback segment is completely offline-meaning that the status in DBA_ROLLBACK_SEGS and V$ROLLSTAT is OFFLINE-you can remove it. To do this, issue the command DROP ROLLBACK SEGMENT segment_name If you need to reduce a rollback segment to its optimal size, you can just wait until this occurs automatically. However, if the segment is taking up space that might be needed by other segments, you can manually cause the segment to shrink by executing the command ALTER ROLLBACK SEGMENT segment_name SHRINK [TO integer[K|M]] As soon as you do this, the rollback segment will shrink. If it doesn't shrink to the desired size as specified in the command, or to OPTIMAL if you didn't specify a size, you may need to re-issue the command later. Some extents may still contain active transactions and so can't be dropped. There's also a chance that your command and the SMON background process were both trying to shrink the rollback segment concurrently. To do this, they must both store some rollback information in the segment themselves, and so may be interfering with the extents that each of them are trying to drop.

Working with Temporary Segments

A temporary segment is a database object that stores information for a server process that can't fit all the data it needs into memory during a sort, a hash join, or other related activities. If it resides in a regular tablespace-that is, one of type PERMANENT-the segment will be available to the server for as long as it's needed. At the end of this time, the segment will be dropped and the blocks it was occupying will be returned to the tablespace for use by another segment. Any number of server processes can be using temporary segments at one time, each managing its own segment and space allocation. If a temporary segment exists in a tablespace defined as a TEMPORARY type, it's not dropped when the initial server process is done with it. Instead, a list of its extents is maintained in the data dictionary, and any other process needing temporary space is allocated one or more of its extents. As with other segments, if there's insufficient space to meet the demand at any given time, more extents are added automatically. The new extents are added to the list of available extents when the processes using them are done. In PERMANENT or TEMPORARY tablespaces, temporary segments obtain their storage information from the default storage defined for the tablespace. The user has no opportunity to set extent sizes, maximum extent counts, and so on. Chapter 5 covers the details of setting up tablespaces with default storage sizes.

Sizing Your Temporary Tablespaces

Temporary tablespaces are difficult to plan for when you first build a database. You don't know just how many SQL statements that might need to use the space are likely to be executing concurrently. You probably won't know how much temporary space any one of them will need. A rule of thumb you might consider is to make the tablespace about half as big as the size of your largest table. This will work against the smaller tables to fit into the tablespace concurrently. However, this may not be sufficient space for work on the largest table to complete, particularly if other work is using the space simultaneously. As your database is used, you can examine performance statistics to see just how many times your applications need temporary segments. Look at the values in the V$SYSSTAT view for the row where the NAME column value is sorts (disk). The number in the VALUE column is the number of sorts since the instance startup that required space in a temporary segment. If you see that the frequency is high enough for multiple such sorts to be occurring simultaneously, you may want to add to your tablespace size. Disk sorts versus memory sorts If you query the V$SYSSTAT view with the statement SELECT * FROM v$sysstat WHERE name LIKE '%sorts%';, you see three rows of data: one for sorts done entirely in memory, one for sorts requiring use of disks (temporary segments), and one showing the total number of rows involved in both types of sorts. If the number of disk sorts is relatively high in comparison to the memory sorts, you may need to tune the memory area provided for sorting, as discussed in Chapter 18. The number of rows sorted shown in V$SYSSTAT may help you determine whether a few relatively large sorts are requiring the temporary segment space, or if a lot of smaller sorts are just spilling over from the sort memory area. Because the row count is accumulated across memory and disk sorts, however, it can be difficult
http://www.informit.com/content/0789716534/element_006.shtml (16 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

to tell how many rows are associated with each type. Also, a sort involving lots of rows may not be as memory intensive as a sort of far fewer, but much longer, rows. To take advantage of this statistic, you may have to monitor V$SYSSTAT on a sort-by-sort basis, with some knowledge of the nature of each sort being recorded in this dynamic performance view. Of course, you may hear from your users if they run out of temporary space, because their applications will fail with an error if they can't acquire sufficient temporary space to complete. You should be careful, however, not to confuse some errors with lack of space in the temporary tablespace: q A temporary segment may reach its MAXEXTENTS limit and not be able to extend any further, even though the tablespace still has room. q Certain DML statements use temporary segments inside a standard tablespace to build the extents for a new or changed segment. If the DML statement can't find the required space, it fails with an error such as ORA-01652 unable to extend temp segment by number in tablespace name. Be sure to check that the named tablespace is really your temporary tablespace before you rush off and try to increase its size; it could be one of your user data tablespaces that's out of room.

Setting Storage Options for Your Temporary Tablespaces

As discussed earlier, temporary segments always build their extents based on the default storage values associated with their tablespace definitions. It's therefore critical that you build your temporary tablespaces with appropriate values in the default storage clause. To understand what storage values are appropriate for temporary segments, you should think about what's being placed into these segments. It's the data that's being swapped out of the space in memory reserved for the type of activity that might need temporary space-either sorts or hash joins. Generally, the space for sorts is the smaller of these two, and the hash join space should be an integer multiple of the sort space size. Therefore, it makes the most sense to build extents that can hold at least the full amount of data likely to be flushed from memory during a sort. If the extent is exactly the same size as the sort space, each set of data flushed to disk will need just one extent. If the extent is larger than the sort space but not a multiple of it, every other write or so from disk will probably need to write into two extents, which isn't as efficient as writing to a single extent. The problem is reversed after the sort when the data has to read back from disk. Some reads will need to skip from one extent to another. If the extent is an integer multiple of the sort size, each set of data flushed from memory will fit into some part of a single extent. However, the last set of data flushed may not fill the extent being used, and the additional space will be wasted until the subsequent processing completes and the entire extent is released. It's almost not worth making each extent in a temporary segment the same as the SORT_AREA_SIZE initialization parameter. If the sort requires more space than this parameter allocates, it will need to write at least two sets of data into its temporary segment-the first set that initially fills up the sort space, and the balance of the sorted data. If no balance were left over to sort, the temporary segment wouldn't have been needed. If the extent size had been double the sort space (2 * SORT_AREA_SIZE), only one extent would have been needed. Use large extent sizes for temporary segments For sorts that almost fit into memory, such as one that uses less than twice the current sort area size, you may well find yourself tuning memory to hold the sort data completely, leaving only the very large sorts in need of temporary space. Such large sorts may need to write a number of sort runs to disk, maybe in the tens or twenties. It makes sense, therefore, to make each extent sufficiently large to hold these multiple runs, thus avoiding the over-head of having to find the required extent space more than once. Your temporary extent sizes may well be 10 to 100 times as large as your SORT_AREA_SIZE, depending on your situation and database use. You may also want to set MAXEXTENTS based on the type of tablespace you're using. If it's of type PERMANENT, you should ensure that each concurrent disk sort can grow its temporary segment to a size that will let the other segments reach the same size. For example, in a 200MB tablespace where you expect five concurrent disk sorts, each temporary segment should be allowed to grow to 40MB. If your extent size is 2MB, MAXEXTENTS would need to be 20. If you're using a TEMPORARY type tablespace, there will probably only be one temporary segment. You can ensure this by building it yourself right after you create your temporary tablespace by using the method described in Chapter 5 You may not want to do this if you're using Oracle Parallel Server, but we'll ignore that case here. If you build the temporary segment yourself, you'll know exactly how many extents can be contained by the tablespace, so just set the tablespace default to that number. Of course, if your tablespace is built with
http://www.informit.com/content/0789716534/element_006.shtml (17 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Managing Redo ... Rollback Segments, and Temporary Segments From: Using Oracle8

files that can autoextend, you should make the MAXEXTENTS value larger. SEE ALSO Building tablespaces specifically for temporary segments,

Managing Your Temporary Segments

Due to their nature, you don't have much control-hence, little management responsibility-for your temporary segments. A couple of issues might affect you: you may run out of space in your temporary tablespace and you may decide to change the value of your SORT_AREA_SIZE parameter. As with any other tablespace, if you run out of space in your temporary tablespace, you can add a new data file or extend an existing data file, assuming that you have the necessary disk space. After you do this, you may want to look at the MAXEXTENTS value for the tablespace to see whether it can be set higher to take advantage of the additional space. Should you change the value of the sort space size by altering the parameter SORT_AREA_SIZE, your temporary extents may no longer align with the I/O to and from the sort space. If you're using a PERMANENT type of tablespace, just changing the default storage values for INITIAL and NEXT to the new value will be sufficient to ensure that the next temporary segments will be appropriately sized. If you have a TEMPORARY type tablespace, however, you'll need to drop the segment from it to free up the space to build one with new extent sizes. To drop the segment(s) from a TEMPORARY tablespace, you need to alter it to be a PERMANENT tablespace. This will cause its contents to be treated as any other temporary segment so they will be dropped automatically. Before a replacement segment is built, you should ensure that the default storage settings are changed to reflect the new extent size you require, as you would do for temporary segments in a PERMANENT tablespace. Therefore, before altering the tablespace back to TEMPORARY status, make sure that you alter the tablespace default storage settings. As when creating a new temporary tablespace of type TEMPORARY, you may want to prebuild the temporary segment to its maximum size as soon as the status is converted. The command to alter the tablespace type is Converting a tablespace to TEMPORARY requires that all other segments be removed from it Anytime a tablespace is of type PERMANENT, other types of segments in addition to temporary segments can be created in it. To switch it to a TEMPORARY type, all such segments must be removed. This is true whether the tablespace was originally created to hold different types of segments or if it was made PERMANENT for only a short time while the storage values were being changed. ALTER TABLESPACE tablespace_name PERMANENT|TEMPORARY SEE ALSO More information about setting up temporary tablespaces and their storage options, Initialization parameters associated with sorts, < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_006.shtml (18 of 18) [26.05.2000 16:47:21]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

You are here : Home : Using Oracle8

Adding Segments for Tables

Exact Phrase All Words Search Tips

< Back

Contents

Next >

Save to MyInformIT

From: Using Oracle8 Author: David Austin Publisher: Que More Information

Table Structure
r r r r r

Choosing a Column Datatype and Length Character Data Numeric Data Date Data Binary Data

q q

Sizing and Locating Tables Setting Storage Parameters

r r r r r

INITIAL NEXT PCTINCREASE MINEXTENTS MAXEXTENTS Creating Tables for Updates Creating Tables with High Delete Activity Creating Tables for Multiple Concurrent Transactions

Setting Space Utilization Parameters

r r r

q q q q q

Building Tables from Existing Tables Monitoring Table Growth Managing Extent Allocation Removing Unused Space Using Views to Prebuild Queries
r r r r r r r r r

Changing Column Names with Views Dropping Columns with Views Hiding Data with Views Hiding Complicated Queries Accessing Remote Databases Transparently with Views Creating and Handling Invalid Views Dropping and Modifying Views Updating Data Through Views View Consistency

http://www.informit.com/content/0789716534/element_007.shtml (1 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

Table Structure
Tables are the most common structures in almost all relational databases. They consist of rows (also known as tuples or instances in the worlds of relational theory and modeling, respectively) and columns (attributes). When queried by Oracle's SQL*Plus tool, they're displayed in a table format, with the column names becoming the headings. Such a display gives the illusion that the data is stored in the database just the way it appears onscreen: q The columns are lined up under column headings. q The data in each column is the same width from row to row. q Numeric data is aligned to the right of the field, other data to the left. q A fixed number of records fits on each page. q Each field is separated from its neighbors by a fixed number of bytes. A query against a table in SQL*Plus could result in the following output: CUST_NUMBER COMPANY_NAME PHONE_NUMBER LAST_ORDER ----------- ----------------------- ------------ ---------100 All Occasion Gifts 321-099-8642 12-MAR-96 103 Best of the Best 321-808-9753 05-MAY-98 110 Magnificent Mark's 322-771-3524 11-DEC-97 111 Halloween All Year 321-998-3623 25-FEB-98 ... Helping SQL*Plus format queries SQL*Plus may not always produce query output just the way you want to see it. For example, a column that's only one character wide will have a one-character column heading by default. Most users won't find the first character of the column name sufficient for identifying the column's contents. Similarly, some columns may be defined to hold many more characters than you need to see when casually querying the table. These longer columns can cause each row to wrap over multiple output lines, making it difficult to read the results. SQL*Plus provides formatting commands to help you produce output that has meaningful column names and column widths. You can even use its advanced features to create subtotals, grand totals, page titles and footers, and other standard reporting features. These are all covered in the Oracle8 SQL*Plus User's Guide. Although this formatting by SQL*Plus is very convenient and emphasizes the notion that relational databases store data in the form of two-dimensional tables, it doesn't really represent the true internal structure of the database tables. Inside Oracle's data files, the rows are stored very efficiently on Oracle database blocks, leaving very little free space (unless it's believed to be needed) and with little regard for how the data will look if displayed onscreen or in a report. The blocks themselves are stored in a data segment consisting of one or more extents. A very simple table will have a single extent, the first block containing its header information and the other blocks storing the rows themselves. A larger table may contain many extents, and a very large table may have additional header blocks to support the more complex structure. The data dictionary maintains the table definition, along with the storage information and other related object definitions, such as views, indexes, privileges, and constraints, some of which can be created with the table itself. (Views are covered at the end of this chapter, indexes in Chapter 8 "Adding Segments for Different Types of Indexes," privileges in Chapter 10, "Controlling User Access with Privileges," and constraints in Chapter 17, "Using Constraints to Improve Your Application Performance.") The simplest version of the CREATE TABLE command names the table and identifies a single column by name and type of data it will hold. Its syntax is as follows: CREATE TABLE table_name (column_name datatype); Another trick to create meaningful column names It's becoming more and more common to prefix the field names with letters that indicate which datatype is stored in the field-for example, dtmStartDate indicates a datatype for the field. You should, of course, choose a name for the table that's meaningful to you and your user community.
http://www.informit.com/content/0789716534/element_007.shtml (2 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

Similarly, the column name should provide some useful information about the purpose of the data it will hold.

Choosing a Column Datatype and Length

Before building a table, you should know what type of information each column will need to hold. By knowing this, you can select an appropriate datatype and, possibly, a length. In some cases, you have a choice of datatypes that could be used for a particular column.

Character Data
You can store freeform character data in a number of formats. Table 7.1 shows the relateddatatypes and characteristics. Table 7.1 Definitions of Oracle datatypes Datatype: BFILE BLOB CHAR CLOB DATE LONG LONG RAW NCHAR NCLOB Length: 4GB 4GB 2,000 4GB 7 2GB 2GB 2,000 4GB Max Preferred Uses: Binary data stored outside the database, allowing fast byte stream reads and writes Variable-length binary objects Short fields or fields that need fixed-length character comparisons Variable-length, single-byte character fields exceeding 2GB Dates and times Variable-length character fields that exceed 4,000 bytes Variable-length, uninterpreted binary data Multibyte characters in short fields or fields that need fixed-length character comparisons Variable-length, multibyte character fields exceeding 2GB; support only one character width per field Number, having precision of 1 to 38 digits and scale of -84 to 127 Variable-length fields that store single or multibyte characters that don't need fixed-length comparisons Variable-length, uninterpreted binary data Extended row Ids Variable-length fields that don't need fixed-length comparisons Variable-length fields that don't need fixed-length comparisons Notes: 1 1 2 1 3 1,4 1,4 2,5 1

NUMBER NVARCHAR2

38 4,000

RAW ROWID VARCHAR VARCHAR2

2,000 10 4,000 4,000

4,6 6,7 6,8

1. There's no default length and no mechanism to define a maximum length. 2. Trailing blanks are stored in the database, possibly wasting space for variable-length data. Default length is 1 character; to provide a maximum field length, add the required length in parentheses following the datatype keyword. 3. Dates are always stored with seven components: century, year, month, day, hour, minute, and second. They can range from January 1, 4712 BC to December 31, 4712 AD. 4. Supported for Oracle7 compliance and may not continue to be supported; large objects (LOBs) are preferred. 5. The maximum length is the maximum number of bytes. For multibyte characters, the total number of characters will be less, depending on the number of bytes per character. 6. There's no default length. You must always supply your own maximum length value in parentheses following the datatype keyword. 7. Will stay compliant with ANSII standard definition for variable-length character fields.

http://www.informit.com/content/0789716534/element_007.shtml (3 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

8. Will stay compliant with the definition from Oracle7. Internally, with the exception of the CHAR and DATE datatypes, Oracle stores only the characters provided by the application in the character fields. If you specify a maximum length (when allowed) or use a predefined type at its maximum length, you don't waste any storage space when your records have fewer characters than the field can hold. The CHAR datatype, however, always adds trailing blanks (if needed) when the supplied data is less than the defined field length. The DATE datatype always uses 7 bytes, one for each date/time component, applying a default specific to each missing component. CHAR Versus VARCHAR2 Besides the storage differences-before being written to the database, CHAR fields are always blank-padded to the full defined length, whereas VARCHAR and VARCHAR2 fields are never padded automatically-the two types of fields sort differently and compare differently. CHAR fields are sorted and compared using their full (padded) length while the variable character fields are sorted and compared on just the characters included in the string. Space management with CHAR and VARCHAR2 Some people prefer to use CHAR rather than VARCHAR2 datatypes to reduce the likeli-hood that rows will grow in length when updates increase the number of characters in a field. Such growth can cause the row to become too long to fit into the block. However, Oracle provides a PCTFREE parameter to allow for row growth. I prefer to use PCTFREE to manage space rather than force all character fields to be padded blank characters, which I consider to be wasted space. See the "Creating Tables for Updates" in this chapter for details on the PCTFREE parameter. A simple test is shown in the following few statements: CREATE TABLE test_padding (fixed_col CHAR(5), var_col VARCHAR2(5)); INSERT INTO test_padding VALUES ('A','A'); INSERT INTO test_padding VALUES ('ABCDE','ABCDE'); SELECT * FROM test_padding WHERE fixed_col = var_col; FIXED_COL VAR_COL --------- ------ABCDE ABCDE Only the row where all five characters have been filled in by the VALUES clause is displayed. The row with the single letters doesn't show the two columns having equal values because the FIXED_COL column is comparing all five characters, including trailing blanks, to the single character from the VAR_COL column.

Numeric Data
Numbers are stored by using the NUMBER datatype. By default, a number field can contain up to 38 digits of precision along with, optionally, a decimal point and a sign. Positive and negative numbers have a magnitude of 1.0 10-130 to 9.99 10125. A number can also have a value of 0, of course. To restrict the magnitude of a number (the number of digits to the left of the decimal point) and its precision (the number of digits to the right of the decimal point), enclose the required value(s) inside parentheses following the NUMBER keyword. If you include only the magnitude, you actually define an integer. Any numbers with decimal values are rounded to the nearest integer before being stored. For example, NUMBER(3) allows numbers in the range of -999 to +999, and an inserted value of 10.65 is stored as 11. If you provide a precision and scale, you can store a number with as many digits as provided by the precision, but only precision-scale digits before the decimal point. Table 7.2 shows some examples of numbers that you can and can't store in a column defined as NUMBER(5,2). Using a negative precision value If you use a negative value in the precision field of a number column's length definition, the numbers will be rounded up to that power of 10 before being stored. For example, a column defined as NUMBER(10,-2) will take your input and round up to the nearest 100 (10 to the power of 2), so a value of 123,456 would be stored as 123,500. How Oracle stores numbers

http://www.informit.com/content/0789716534/element_007.shtml (4 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

Oracle stores all numbers, regardless of the definition, by using a mantissa and exponent component. The digits of the mantissa are compressed two digits per byte, so the actual space required to a store a number depends on the number of significant digits provided, regardless of the column definition. Table 7.2 Valid and invalid numbers for a column defined as NUMBER(5,2) Valid Numbers: 0 1 12.3 -12 -123.45 123.456 -12.345 Valid Numbers: 123.4567890123456789 12345 -1234.1 Stored As: 0 1 12.3 -12 -123.45 123.46 (rounded to 2 decimal digits) -12.35 (rounded to 2 decimal digits) Stored As: 123.46 (rounded to 2 decimal digits) Invalid; exceeds precision (5-2 digits before decimal) Invalid; exceeds precision (5-2 digits before decimal)

Handling the year 2000 problem Oracle has always stored both the century and the year for any date value in the database. To help distinguish dates in the 20th and 21st centuries when you provide only the last two digits of the year for the TO_DATE function, Oracle provides the RR date format mask. In a statement that stores a date by using the function TO_DATE('12/12/03',' DD/MM/RR'), the stored date will have the current century if the current year's last two digits are less than 50, and will have the next century if the current year's last two digits are 50 or greater. Full details of the RR format are given in the Oracle8 Server SQL Reference manual.

Date Data
Use the DATE format to store date fields or time information. Oracle has a single 7-byte internal format for all dates, and a 1-byte interval for each century, year, month, day, hour, minute, and second. Depending on the format your applications use to store dates, some fields may be left to default. For the time fields, the defaults result in a time of midnight. For century, either the current year (taken from the operating system date setting) is used, or a choice of 1900 or 2000, depending on the year value. The RR format mask causes the latter behavior, with a year in the range 50 to 99 resulting in a value of 19 for the century, and a year in the range 00 to 49 resulting in a value of 20 for the century. Default date formats Oracle uses a format mask when dealing with dates so that each of the seven components of the combined date/time fields can be uniquely identified. The database runs with a default date mask dependent on the setting of the initialization parameters NLS_TERRITORY and NLS_DATE_FORMAT. Generally, these hide the time component so that it defaults to midnight, unless the application or individual statement decides to override the mask and provide its own values for one or more of the time fields. When you're using date arithmetic and date functions, the time component may not be obvious to you or your users and can cause apparent problems. Entering just the time component causes the date portion to be derived from the current operating system date, which uses the RR format mask process as described for the century as well as the current year and month, and defaults the day to the first day of the month. Oracle can perform extensive date field operations, including date comparisons and date arithmetic. If you need to manipulate dates, check the Oracle8 Server SQL manual for detailed descriptions of the available date operators and functions.

http://www.informit.com/content/0789716534/element_007.shtml (5 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

One common problem occurs when you use the SYSDATE function to supply the current date when inserting a new record into the database. This would seem to be straightforward, allowing a query such as to select all orders placed on October 10, 1997, assuming that the date is provided in the correct format. SELECT * FROM orders WHERE order_date = '10-OCT-97' However, because the SYSDATE function by default always inserts the current time as well as the current date (whereas the query provides only the date, meaning that midnight on October 10 is being selected), there will be no matching records in the ORDERS table. The solution would be to store the data as if the time were midnight by applying the TRUNC function to the SYSDATE on insert.

Binary Data
Binary data is stored without any interpretation of embedded characters. For compatibility with Oracle7, the RAW and LONG RAW datatypes are still usable. However, the LONG RAW datatype is being deprecated, meaning that it's gradually becoming unsupported. Oracle8 offers the BLOB and BFILE datatypes to store binary data; these can be used in place of the RAW and LONG RAW datatypes. SEE ALSO For details on large objects (LOBs), The only internal difference between RAW and LONG RAW is the maximum number of bytes they can store. RAW has a maximum length of 2,000 bytes, and you must define the maximum length you need as part of the column definition, even if you need all 2,000 bytes. The LONG RAW datatype can hold a maximum of 2GB. You can't limit this size as you can with RAW column, but, as with variable-character fields, Oracle stores only the characters you supply in RAW and LONG RAW fields, regardless of the maximum possible length. By using some of the datatypes discussed above, you could create a multicolumn table, SAMPLE1, as follows: CREATE TABLE sample1 ( sample_id NUMBER(10), sample_name VARCHAR2(35), owner_id NUMBER(4), collection_date DATE, donor_gender CHAR(1), sample_image BLOB); The new syntax isn't very complicated Compared with the CREATE TABLE command included at the beginning of this chapter, the only other new syntax introduced in the listing, in addition to the datatypes, is the comma that separates each column definition. Tables defined with a variable-length datatype in one or more columns-that is, any datatype other than CHAR or DATE-may need some special consideration when they're created if these columns are likely to be updated during the lifetime of any given row. This is because Oracle packs the data into table blocks as tightly as possible. This tends to result in very little, if any, space being left on the block if a row grows in length due to an update that adds more bytes to an existing field. Before creating a table in which you anticipate updates being made to variable-length columns, read the later section "Setting Space Utilization Parameters" to see how to avoid some of the problems this can cause.

Sizing and Locating Tables

Appendix A of the Oracle8 Server Administrator's Guide provides a detailed description of calculations you can use to compute how big a particular table will be, based on the column definitions, the average record size, and the expected number of rows. Before the most recent releases of Oracle, a table could contain only a limited number of extents, so it was important to know how big the table would be before creating it in order to provide sufficient extents space. With Oracle8, you can have unlimited extents in a table. Consequently, the only reason to be concerned with table size is to determine if you have sufficient disk space to store it. Although I won't try to dissuade you from using Oracle's provided sizing calculations, I want to suggest an alternative approach to predicting table size. One key to either approach is to know the average row size. This generally requires that you find or generate some valid sample data. If you have such data, I recommend that you simply load it into your table, measure how much space is used, and then extrapolate the final size based on the ratio of the number of rows in your sample data to the total number of rows you expect the table to contain.

http://www.informit.com/content/0789716534/element_007.shtml (6 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

Use sample data to predict table size 1. Put your sample data into a flat file. 2. Create a SQL*Loader control file and run SQL*Loader to load the data. (See Chapter 25, "Using SQL*Loader and Export/Import," for details on SQL*Loader.) 3. Execute the following command to collect current storage information: ANALYZE TABLE table_name COMPUTE STATISTICS 4. Execute the following query to find the number blocks now storing data: SELECT blocks FROM user_tables WHERE table_name = 'table_name' 5. Compute the total number of blocks required to store the full table by using this formula: BLOCKS X(total number of rows) / (number of rows in sample) You can use the following SQL*Plus script to perform these steps after you load your sample rows: SET VERIFY OFF ANALYZE TABLE &table_name COMPUTE STATISTICS / SELECT blocks * &total_row_count / num_rows AS "Total blocks needed" FROM user_tables WHERE table_name = UPPER('&&table_name') / After you determine your table's maximum size, you can identify an appropriate tablespace in which to store it. Your choice should be based on the factors discussed in Chapter 5 "Managing Your Database Space," concerning tablespace usage. These include a recommendation to use a limited number of different extent sizes, such as small, medium, large, and huge, for all objects in a given tablespace. You should be able to determine from its maximum size which category of extent sizes would best suit it, assuming that you follow our recommendations. For a very large table, the largest extent size is usually preferable, although if the table is going to grow very slowly, you may want to use smaller extents so that you can conserve disk space in the interim. Other factors in deciding on a tablespace include the frequency of backup, the likelihood of dropping or truncating the table in the future, and which other segments exist in the candidate tablespaces. The latter might influence your decision when you consider what else the application might need to have access to, besides the new table, if a data file in the tablespace should become unusable. Permissions when creating a table To create a table successfully, you must have the necessary privileges and permissions, including the CREATE TABLE privilege and the right to use space in the named tablespace. See Chapter 9 "Creating and Managing User Accounts," and Chapter 10 for more information on these topics. When you've determined which tablespace to use, you should add the tablespace name to the CREATE TABLE statement. By using the SAMPLE1 table-creation script shown in the preceding command, let's put the table into the SAMPLE_DATA tablespace: CREATE TABLE sample1 ( sample_id NUMBER(10), sample_name VARCHAR2(35), owner_id NUMBER(4), collection_date DATE, donor_gender CHAR(1), sample_image BLOB) TABLESPACE sample_data / Moves sample1 table into sample.data tablespace

http://www.informit.com/content/0789716534/element_007.shtml (7 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

Of course, if the person creating the table has the SAMPLE_DATA tablespace as his or her default tablespace, the TABLESPACE clause isn't needed. If you include it, you guarantee that the table will be created in the desired tablespace no matter who runs the script.

Setting Storage Parameters

You need to concern yourself with a table's storage parameters only if you haven't designed your database's tablespaces according to the guidelines in Chapter 5 These guidelines, along with the correct tablespace selection discussed in the previous section, should allow the table to use the default storage information defined for the tablespace. However, we'll look at each storage option in turn for those of you who may need to consider overriding your defaults. You include these options with the required values in a STORAGE clause as part of the CREATE TABLE or ALTER TABLE commands. Overriding the defaults If you have a database that's not designed as rigorously as discussed earlier, you may need to override one or all of the storage parameters. You can find the current default settings for the storage options in a tablespace by querying the DBA_TABLESPACES data dictionary view.

INITIAL
This parameter sets the size, in bytes, of the first extent built for the table. Some possible criteria for choosing a size include the following: q For fast table scans, the extent should hold the entire table. q For fast parallel table scans, the extent should hold 1/xth of the table, and the rest of the table should be placed in other (x-1) equal-sized extents on different disks. q To load the table using SQL*Loader in parallel, direct path, the extent should be as small as possible because it won't be used. q Fit as much of the table as possible into the largest piece of free space available. This is particularly useful when the tablespace has lots of free space but only small amounts of it are contiguous.

NEXT
This parameter sets the size, in bytes, of the second extent. Some possible criteria for choosing a size include the following: q For fast table scans, the extent should hold all the rows not stored in the first extent. q For fast parallel table scans, the next extent should be the same size as the initial and all subsequent extents, and each extent should be stored on a different disk. q To load the table with SQL*Loader in parallel, direct path, the extent should be large enough to hold all the rows from one parallel loader session. q Fit as much of the table as possible that doesn't fit into the initial extent into the largest piece of remaining free space. This is particularly useful when the tablespace has lots of free space but only small amounts of it are contiguous.

PCTINCREASE
This parameter defines a multiplier to compute the size of the next extent to be created. It's applied to the third, and every subsequent, extent. If you set it to zero (0), each extent will be the same size as defined by NEXT; if you set it to 100, each subsequent extent will double in size. A value of zero is generally to be preferred. You may want to use a non-zero value if you don't know how much your table will grow, so that each extent will be larger than the previous one. This should eventually result in a sufficiently large extent to hold the remainder of the table. PCTINCREASE options

http://www.informit.com/content/0789716534/element_007.shtml (8 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

Oracle allows large values of PCTINCREASE to reduce the number of additional extents that might be needed if a table's size was seriously under-estimated when it was first created. In earlier releases, this feature was essential because the number of extents that could be added to an existing table was a finite, limited number. With the UNLIMITED option now available, the only drawback to having many extents is the overhead associated with adding each new one. In general, I recommend leaving the value at zero whenever the table must share a table-space with at least one other segment, to preserve uniform extent sizes. In other cases, you should set it to a reasonable value so that it doesn't begin requiring extents significantly larger than the available disk space. The drawbacks to this include the following: q An attempt to create an extent larger than can be held by any data file in the tablespace. q Irregular extent sizes, leading to irregular free extents if the table is dropped or truncated. q A final extent much larger than required for the number of rows it needs to hold. q Less predictable space consumption, particularly if there are many tables so defined.

MINEXTENTS
This parameter sets the number of extents built by the CREATE TABLE command. The sizes of the extents are determined by the values for INITIAL, NEXT, and PCTINCREASE. There are some possible reasons for creating only one extent initially: q You expect the table to fit into a single extent. q Additional extents will be built by SQL*Loader in parallel, direct mode. q You'll add extents manually to fit them into differently sized free extents. q You'll add extents manually to place them on different disks. There are some possible reasons for creating multiple extents initially: q You have lots of free extents with equal, or nearly equal, sizes, but none large enough to hold the whole table. q Your tablespace is built with many data files, and you want Oracle to spread the extents evenly among them.

MAXEXTENTS
This parameter sets the maximum number of extents the table will be allowed to use. You don't usually need to worry about this value initially because it can be changed later. Of course, you should be prepared to monitor the use of the space as the number of extents in a table approaches this value, no matter how you've set it. In some situations, however, you may need to set a specific value of MAXEXTENTS, including q When you have limited disk space. q When you have fragmented free space in such a way that NEXT and PCTINCREASE can't be set to reasonable values until the next extent is needed. q In a parallel server environment, when you're manually assigning extents to specific instances (see the Oracle8 Server Parallel Server Administration manual for more details on this topic). Additional storage options that don't affect extent sizes You can use other keywords in the STORAGE clause of the CREATE TABLE command-FREELISTS, FREELIST GROUPS, and BUFFER POOL. However, these values can't be set at the tablespace level and don't affect the allocation of table extents. The impact of these storage options is discussed in n other sections of this book. SEE ALSO To see how to use the BUFFER POOL keyword, For more on free lists and free list groups, The buffer pool and table options to use it effectively are covered on We end this section by showing the additional lines added to the CREATE TABLE sample1 script to include a STORAGE clause: CREATE TABLE sample1 ( sample_id sample_name

NUMBER(10), VARCHAR2(35),

http://www.informit.com/content/0789716534/element_007.shtml (9 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

owner_id collection_date donor_gender sample_image TABLESPACE sample_data

NUMBER(4), DATE, CHAR(1), BLOB)

The STORAGE clause

Setting Space Utilization Parameters

Each block in an Oracle table can hold as many rows as will fit into the block. When the block is full, Oracle removes it from the list of blocks into which new rows can be inserted. At some point in the future, if enough rows are deleted from the block, it may be added back onto the list of available blocks so that more new rows can be added to it, using the space freed up by the dropped rows. Some special parameters associated with the table's definition, known as space utilization parameters, influence these events. In particular, they control just how full a block becomes before it's moved off the list of available blocks (known as the free list), how much space must be made available before it's moved back onto the free list again, and how much space is reserved for multiple transactions to access the block concurrently. These settings affect how updates are managed, how much space might be going to waste in a table, and how much transaction concurrency can occur on a block.

Creating Tables for Updates

Unless your table contains all DATE and CHAR columns, any update to a record can cause that record to grow or shrink in overall length. This is because Oracle stores only the bytes that contain information in its variable length fields. A new value, if it has more or less data, changes the amount of storage required. Updating a record with smaller field values doesn't cause any problems, unless this happens repeatedly and the block becomes almost empty, thus wasting space. However, if you add to a field's length, you may run out of space on the block. Even if this doesn't happen the first time you perform such an update, it may occur if you continue to increase the lengths of different columns or rows. If you run out of space, Oracle will move the row to another block in the table through a process known as migration. Although this may sound like a benign solution, it does have performance ramifications. A migrated row leaves behind a forwarding address (pointer) so that it can still be found by an index lookup or by an ongoing query. Subsequent access to that row results in a probe of the row's original home block, which finds the forwarding address, and then a probe of the block where it now resides, known as an "overflow block." Rarely will the original block and overflow block be contiguous blocks in the table, so the disk retrieval for such a row will be slow. Migrated rows, particularly if they are numerous, can affect the overall database performance. To help you avoid massive row migration, Oracle lets you reserve space on a block into which the data can expand. By default, this space is 10 percent of the data area of the block. When a change to the block leaves less free space than is reserved for row expansion, the block is taken off the free list and no further inserts will occur. For some tables, the 10 percent default may be perfectly adequate. For other tables, 10 percent may be a completely inadequate amount of free space or far more than is needed. The space is reserved with the PCTFREE parameter, and you should determine a good value for it as best you can before you build any production table. As soon as a table is built you can change this value, but only blocks that aren't being used will adopt the new value and reserve the desired amount of free space. If you know two pieces of information about the data being stored in the table, you can use the following formula to compute a good value for PCTFREE. The information you need is the average length of the rows when they're first inserted and the average length of the rows when they're at their maximum length. In the following formula, the terms avg_insert_length and max_length refer to these values, respectively:

http://www.informit.com/content/0789716534/element_007.shtml (10 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

(max_length - avg_insert_length) PCTFREE = 100 * --------------------------------------(max_length) In determining the average row lengths, you need to consider only the number of bytes of data per row, not the internal overhead associated with stored rows and fields. If you use just data lengths, the result will be slightly higher and have a built-in margin of error. The size of this error varies depending on the block size, number of rows in the block, and number of columns per row. For a table with 10 columns and a database block size of 4KB, if 10 rows fit into the block, this margin of error will be just over 3 percent. Example of computing the PCTFREE value If a table has a row with an average length of 153 bytes when it's initially inserted, and it grows by an average of 27 bytes over the course of its time in the table, the average maximum length of a row is 180 bytes. By using these two values in the formula PCTFREE = 100 * (max_length -avg_ insert_length) / (max_length), we find that this table should be created with PCTFREE = 100x(180-153) / 180 = 100x27 / 180 = 100x3/20 = 15. For rows that don't change over time, or change only fixed-length fields, the expression (max_length avg_insert_length) reduces to zero, which in turn causes the entire formula to result in zero. If you're really certain that there will be no updates or just updates that change record lengths, you can set PCTFREE equal to zero without concern for row migration problems. If you have a table in which the value of (max_length - avg_insert_length) is negative, you also shouldn't have to worry about migration if you set PCTFREE to zero. However, in such a table, there will be a tendency for the amount of data on each block to become less than is optimal. This will occur when the block gains sufficient empty space, due to record shrinkage, to hold a whole new row. With many blocks in this state, you'll suffer some inefficiency because of this wasted space; more blocks are being taken to store rows than are really needed. To overcome this, you should consider the table in the same category as tables that undergo record deletions over time, and follow the approach to deal with these in the next section.

Creating Tables with High Delete Activity

Over time, tables that have rows deleted or rows that shrink in size can become inefficient. The empty space on the blocks represents additional disk reading and writing that must be done because empty space, rather than data, is being transferred between disk and memory. Oracle provides the space utilization parameter PCTUSED to help you control the amount of empty space allowed to remain on a block. As mentioned earlier, a block is taken off the free list when it's full so that no further attempts are made to insert more rows into it-that is, when it has less empty space than PCTFREE of the block. PCTUSED sets a threshold value at which the amount of free space becomes sufficient for the block to hold one or more new rows, so it can be placed back on the free list. By default, Oracle sets PCTUSED at 40 percent. In other words, a block that has less than 40 percent of its data area occupied by rows will be put back on the free list. You can change the value of PCTUSED at table creation time or anytime thereafter. As with PCTFREE, the impact of a change to PCTUSED may be delayed. A block that already contains less than the new PCTUSED amount of data, unless it's already on the free list, won't be placed there until another change is made to it. What makes a good value for PCTUSED? The first criterion is to set it so that a block goes back on the free list only when there's room to store at least one more new row. In a very volatile table, where rows are frequently added and dropped, it may be worth wasting space on a block until there is room to fit three or four rows. Moving blocks on and off the free list requires some overhead that may not be worth incurring unless more than one row is affected by the change. After you decide how many rows to leave room for before placing a block back on a free list, you can use the following formula to compute a good value for PCTUSED: PCTUSED = 100 - PCTFREE - 100 * row_space / block_space where row_space = avg_insert_length * rows_needed and block_space = DB_BLOCK_SIZE - 90 - INITRANS * 24 PCTFREE is a space-utilization parameter value discussed in the previous section. avg_insert_length is the average number of bytes in a row when it's first inserted. rows_needed is the number of rows you want to be able to fit into the block before returning it to the free list. DB_BLOCK_SIZE is the database block size, set in the parameter file and found by querying the

q q q

http://www.informit.com/content/0789716534/element_007.shtml (11 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

V$PARAMETER table. INITRANS is a space-utilization parameter value discussed in the next section.

Constants used in computing PCTUSED value The constant 90 is an imprecise measure of the space used by Oracle's header information in a table block, but it has proven to be sufficiently accurate for this calculation. The constant 24 is the number of bytes used to store a transaction entry on a typical hardware platform, and should be adequate for this calculation. Following from the example we used to demonstrate the computation for PCTFREE, let's see how this formula would work if we wanted to insert new rows when a block had room for three new rows. In the earlier example, the average length of a row when initially inserted was 153 bytes, and the value for PCTFREE was calculated at 15. Let's use a block size of 4KB and an INITRANS value of 4 to complete the PCTUSED calculation. So we need to compute PCTUSED = 100 - 15 - 100 * (153 * 3) / (4096 - 90 - 4 * 24) 100 is a constant for computing the percentage. 15 is the value for PCTFREE. 153 is the number of bytes required to store an average row when inserted. 3 is the number of rows we need to have room to store before returning the block to the free list. 4096 is the number of bytes on a 4KB block. 90 is a constant representing the number of bytes used for block overhead. 4 is the value of INITRANS. 24 is a constant representing the number of bytes taken by a transaction entry.

q q q q q q q q

This simplifies to the following: PCTUSED = 85 - 100 * 459 / (4006 - 96) = 85 - 100 * 459 / 3910 If we round the quotient 459/3910 (= 0.1173913) up to 0.12, the result becomes the following: PCTUSED = 85 - 100 * 0.12 = 85 - 12 = 73 The second consideration is how much space you can afford to spare. The lower the PCTUSED value you use, the more empty space will accumulate on a block before it's recycled onto a free list for more data to be added. In very large tables, you may not be able to afford to store blocks with more than a minimal amount of free space. In such cases, even though you may cause additional overhead by moving blocks back onto the free list more often than you might think you need from the preceding formula, you may gain some benefits. Not only will you save disk space, but if the table is queried extensively-particularly when using full table scans-you'll need to read less blocks into memory to retrieve the same number of rows.

Creating Tables for Multiple Concurrent Transactions

For a transaction to add, alter, or drop a row from an Oracle table, it must first obtain a lock on that row. It does this by first registering itself on the block where the row will reside or now resides. The registration is made by updating a special area of the block called the transaction entry slot, also known as an intent to lock (itl) slot. If a block contains a lot of rows, it's conceivable that more than one transaction will want to work on the same block at the same time. To do this, each must obtain a transaction slot for its own use. Oracle allows up to 255 transaction slots to be created on a single block, but by default it builds one only when a block is added to a table. When additional slots are needed, the Oracle server process needing the slot has to rearrange the contents of the block to make room for the new slot. If you want to avoid this behavior, you can create a table that will contain more transaction slots on each block as the blocks are added. You can also limit the upper number of slots that can be created, preserving the space for additional row data, albeit at the cost of possibly making users wait for a transaction slot on a very busy block. You control the allocation of transaction slots with the INITRANS and MAXTRANS space utilization parameters. With INITRANS, you set the number of transaction slots that each block acquires by default. With MAXTRANS, you set an upper limit on the total number of such slots that can be assigned to the block. The difference, MAXTRANS minus INITRANS, is the number of slots that can be added dynamically if needed. Usually, there's no real need to change the default value for MAXTRANS. Even if you have hundreds of concurrent transactions working against the same table, they're likely to be working on different blocks simply
http://www.informit.com/content/0789716534/element_007.shtml (12 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

because most blocks don't have room for that many rows. In the rare situation where tens of concurrent transactions all need the same block, they'll probably have to wait for one of the other transactions to release the row-level lock before they can do any work. It's in this case that you might want to set MAXTRANS. Otherwise, each transaction will build itself a transaction slot that it will then occupy idly until it can get to the row it needs. These slots represent wasted space on the block. You might want to change INITRANS, however, if your can predict that more than one transaction will likely need the same block at the same time. By preallocating the necessary number of transaction slots on each block, you'll help the second and subsequent user get to their resources sooner. Each slot requires about 24 bytes, so don't set the value of INITRANS too high. Otherwise, you'll be taking space that could be occupied by row data. Adding space utilization parameters to the example SAMPLE1 table requires further modifications to our table-creation command: CREATE TABLE sample1 ( sample_id sample_name owner_id collection_date donor_gender sample_image TABLESPACE sample_data STORAGE ( INITIAL NEXT PCTINCREASE MAXEXTENTS

NUMBER(10), VARCHAR2(35), NUMBER(4), DATE, CHAR(1), BLOB)

5M 5M 0 50)

Space utilization parameters

Building Tables from Existing Tables

One option you may want to exercise is to build a table from the definition of, or the full or partial contents of, another table. Oracle allows you to do this through an AS SELECT clause to the CREATE TABLE command. The table definition can look just the same as the examples you've seen to this point, except that the datatype isn't included in the column list because the type is inherited from the original table. In fact, if you also want to keep the column names in the new table the same as they are in the original table, you can omit the column list completely. As with the versions of the CREATE TABLE command you've seen listed, you can also include or omit the entire STORAGE clause, or just include it with the required parameters; you can include or exclude any or all space utilization parameters; and you need to include the TABLESPACE clause only if you want the new table to be created somewhere other than your default tablespace. The AS SELECT clause can include any valid query that will retrieve columns and rows to match the new table's definition. The columns named in the SELECT clause must match the column list, if any, for the new table. If the new table doesn't have a column list, all the columns from the original table are used in the new table. The SELECT clause can optionally include a WHERE clause to identify which rows to store in the new table. If you don't want any rows stored, include a WHERE clause that never returns a valid condition, such as WHERE 1 = 2. The following shows three different variations of the CREATE TABLE...AS SELECT statement, each one producing a different table definition from the same base table we have been using, SAMPLE1. REM Create SAMPLE2, an exact copy of SAMPLE1, in tablespace REM SPARE, using default storage and no free space in the REM blocks.

http://www.informit.com/content/0789716534/element_007.shtml (13 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

REM

SAMPLE2 is based on entire SAMPLE1 table

REM REM REM REM REM

Create SAMPLE3, containing just the ID and IMAGE columns, renamed, from SAMPLE1, placing it in the IMAGE tablespace with unlimited l00MB extents and default space utilization parameters.

SAMPLE3 is based on two renamed columns from SAMPLE1 table

REM REM REM REM REM REM

Create SAMPLE4 containing all but the IMAGE column from SAMPLE1, and only selecting records from the past year. Use the DEMOGRAPHIC tablespace with default storage, zero free space, a block reuse threshold of 60 percent, and exactly 5 transaction slots per block.

http://www.informit.com/content/0789716534/element_007.shtml (14 of 21) [26.05.2000 16:47:43]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

SAMPLE4 is based on all but one column from the SAMPLE1 table and includes only a subset of rows

Monitoring Table Growth

Although you try to determine a table's overall size when it's created to evaluate disk requirements, the actual size may well vary from the predicted size. You should monitor the database tables to ensure that they won't run of out space in the course of normal business. Similarly, you may want to check that the space taken by the table's extents is being used efficiently and that there aren't a lot of empty or near-empty blocks. You may also want to confirm that the PCTFREE value is set appropriately by looking for migrated rows. The ANALYZE command collects statistics and stores them in the data dictionary for you. These statistics include the number of blocks used, the amount of unused space per block, the number of empty blocks, and the number of migrated or chained rows. A chained row is one that's simply too large to fit into a single block, and thus will always be spread across multiple blocks. The data dictionary, unfortunately, doesn't distinguish between chained rows and migrated rows, the latter being rows that get longer through the use of UPDATE commands and don't have sufficient space on their block for the growth. If the average row length, also shown in the dictionary, is less than the space available on a block (the block size minus the header and transaction slot space), the rows are most likely migrated, not chained. Row chaining versus migration If a row is too big to fit into a single block, the first part of the row is stored in one block and the rest of the row is stored in one or more overflow blocks. Each part of the row is known as a row piece, and the first row piece is counted as the row's location for any index entry or when it's examined by the ANALYZE command. When a row migrates because it no longer fits into its original block, a pointer is left behind in the original block to identify the row's new location. This pointer is treated as the row's location so that any index entries pointing to the row don't have to be updated. The ANALYZE command treats this pointer as an initial row piece and doesn't distinguish it from a row piece belonging to a chained row. This is why the results of the ANALYZE command don't distinguish between chained and migrated rows. You can use ANALYZE TABLE table_name COMPUTE STATISTICS or ANALYZE TABLE table_name ESTIMATE STATISTICS to collect the statistics stored in the DBA_TABLES (and USER_TABLES) data dictionary views. The former command will always give accurate results; the latter will be a good, but not precise, estimate. The COMPUTE option takes longer as the table grows in size, so you may prefer to estimate statistics for your large tables. You can select the percentage of the table or the number of rows you want to include in the estimate with the SAMPLE clause, using SAMPLE x PERCENT or SAMPLE x ROWS (where x is the percentage or the row count, respectively). When you collect statistics, you may affect how the database performs optimization to determine statement execution plans. If you want to ensure that rule-based optimization is used by default, you should execute the
http://www.informit.com/content/0789716534/element_007.shtml (15 of 21) [26.05.2000 16:47:44]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

ANALYZE TABLE...DELETE STATISTICS command after you examine the statistics. If you want to spend more time reviewing the statistics, you can save the results by executing a CREATE TABLE...AS SELECT command against the data dictionary table. In fact, if you do this after you run the ANALYZE command to collect new statistics, using a different table to store the results each time, you will build a history of the table's growth and data distribution. Once you have saved the statistics into a table, you can go ahead and execute the DELETE STATISTICS option to remove them from the base table definition.

Managing Extent Allocation

There are two reasons to monitor how many extents exist in a table: q You don't want the table to reach its MAXEXTENTS number of extents and thus fail during the execution of a command. q This may be a table where, for whatever reason, you want to add extents manually, and so you maintain MAXEXTENTS at the current number of extents to avoid dynamic allocation. Reasons to add table extents manually There are a number of reasons for manually adding extents to a table, although relying on automatic allocation generally makes your work a lot easier. Some of the more common reasons for manual allocation include over-coming a shortage of space in the tablespace that doesn't allow you to choose a good value for the NEXT extent to be allocated automatically; placing the extent in a data file of your choice, which allows you to spread the storage around different data files and, presumably, different disk drives; allocating the extent to a specific instance's free list groups if you're using the parallel server option; and ensuring that you fit the largest extent possible into the given space. To see the number of extents in table, you can query DBA_EXTENTS for the given table (segment) name. If you use the COUNT(*) value in the SELECT clause, the result will show the exact number of extents owned by the table. If you want to see other information, such as the data files in which the extents are stored, you can query other columns in this table. To allocate an additional extent to a table for which you're using manual allocation, issue an ALTER TABLE...STORAGE (MAXEXTENTS x) command, where x is one more than the current number of extents. You can then add an extent with the ALTER TABLE...ALLOCATE EXTENT command. The following script manually adds the 12th extent to the SAMPLE10 table, using the third data file in the USR_DATA tablespace: ALTER TABLE sample10 STORAGE (MAXEXTENTS 12) / ALTER TABLE sample10 ALLOCATE EXTENT ( FILE 'c:\orant\database\samples\usr_data3.ora') / For tables to which extents are being added automatically, you simply need to ensure that MAXEXTENTS stays larger than the current number of extents. By monitoring the table over time, you should be able to predict how fast extents are being added and increase the MAXEXTENTS value before the current limit is reached. You use the same command that appears at the start of the preceding script to change the extent limit.

Removing Unused Space

There are two types of unused space in a table: q Blocks that have been allocated but never used for data. q Blocks that have become empty or partially empty over time due to row deletions or smaller values updated in variable-length fields. If you have blocks of the first type-that is, blocks that have never been used-and don't expect this space to be needed, you can remove it with the DEALLOCATE option of the ALTER TABLE command: ALTER TABLE table_name DEALLOCATE UNUSED If you expect some but not all allocated space to be needed, you can drop a portion of the extra space by using a further option:

http://www.informit.com/content/0789716534/element_007.shtml (16 of 21) [26.05.2000 16:47:44]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

ALTER TABLE table_name DEALLOCATE UNUSED KEEP integer [K|M] This removes all but integer bytes (or K (kilobytes) or M (megabytes)). To reclaim the other type of free space-space that has been released by DML activity-you can try increasing the PCTUSED value for the table, as discussed earlier. This will allow blocks to be returned to the free list and used for future new rows sooner than they have been. However, if the table is fairly static and not many more changes will be made, the blocks that are already partially empty can't be touched again and won't be returned to the free list. Even if they were, there might not be enough new rows added to fill all the reusable space. In this case, you may have to rebuild the table. You can rebuild a table in a number of ways: q Use the Export/Import utilities explained in Chapter 25.
q

Dump the records into an external file and use SQL*Loader (also explained in Chapter 25) to reload them. If you have room, you can move the records to a temporary table, truncate the original table, and move the records back again, as shown in the following for the SAMPLE10 table: CREATE TABLE temp AS SELECT * FROM sample10 / TRUNCATE TABLE sample10 / INSERT INTO sample10 SELECT * FROM temp / DROP TABLE temp /

Using Views to Prebuild Queries

Although you can create tables from the contents of other tables, as discussed in the previous section, in many cases there's no need to consume the space required by the copy just because you need to see a variation of the original table. A view is built in the same way that the AS SELECT clause is used to build a table from the contents of another table. With a view, rather than the data be copied to a new location, only the definition of required data is stored in the data dictionary. This amounts to storing the SELECT statement or, to put it another way, storing a query definition. Views have many uses. The following sections show you how to build a view to meet a specific need you may have. We use a simple EMPLOYEE table as the basis for the examples in these sections. The following shows the CREATE statement for this table: You can use views without degrading performance Consider using a view whenever you think it could be useful, for any purpose whatsoever. There is very little overhead involved in storing the definition or in executing a statement against the view. CREATE TABLE employee ( id NUMBER(8) CONSTRAINT employee_id_pk PRIMARY KEY, last_name VARCHAR2(35), first_name VARCHAR2(30), middle_initial CHAR(1), department NUMBER(5) CONSTRAINT employee_department_fk REFERENCES department_table, salary NUMBER(10,2), title VARCHAR2(20), phone NUMBER(5) CONSTRAINT employee_phone_fk REFERENCES phone_table, hire_date DATE) / For information on the constraints contained in this table definition, see Chapter 17, "Using Constraints to
http://www.informit.com/content/0789716534/element_007.shtml (17 of 21) [26.05.2000 16:47:44]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

Improve Your Application Performance."

Changing Column Names with Views

If your table has column names that follow a naming convention-such as a corporate standard or one based on the vocabulary of the primary users-the names may not be meaningful to other users of the table. For example, the Human Resource Department may talk about an employee ID number, the Payroll Department may refer to the same number as a "payroll number," and the Project Scheduling System may use the term "task assignee." By using views, the same employee table can be used to provide this number with the preferred name for each group. The following script, based the EMPLOYEE table created in the preceding script, shows a view being created for the Payroll Department to give the ID column the name EMPLOYEE_ID, leaving all other columns with their original names: CREATE VIEW pay_employee ( payroll_number, last_name, first_name, middle_initial, department, salary, title, phone, hire_date) AS SELECT * FROM employee /

Dropping Columns with Views

Once in a while, you may find that a column originally defined in a table is no longer needed. The current version of Oracle8 doesn't allow you to drop such a column. Instead, I recommend that you set the column value to NULL in the entire table, which will free up the storage consumed by the column. This doesn't help users who may not know about the column's existence when they try to use the table. An INSERT statement, for instance, would fail if they didn't include a value for the dropped column. If you create a view that excludes the missing column, the view can now be used in place of the table and the dropped column will no longer be a problem. Updating views may not always be possible If you have views that participate in table joins, your users may not be able to update them or perform other DML commands on them. For detailed information on the rules governing updatable views, see the later section "Updating Data Through Views." To make the column's disappearance even more transparent to the users, you can first rename the table and then use the original table name to name the view that excludes the unwanted column: 1. Issue the following command to prevent table access under the old name: RENAME table_name TO new_table_name; 2. Build the required view with this command: CREATE VIEW table_name (...) AS SELECT ... FROM new_table_name; Make sure that you name all but the unwanted column in the list of column names (shown as in the preceding statement) and use the new name for the table in the FROM clause. 3. Grant the same permissions on the view that existed on the table. The following CREATE VIEW statement will result in a view that apparently removes the PHONE column from the EMPLOYEE table: CREATE VIEW emp AS SELECT id, last_name, first_name, middle_initial, department, salary, title, hire_date

http://www.informit.com/content/0789716534/element_007.shtml (18 of 21) [26.05.2000 16:47:44]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

FROM employee /

Hiding Data with Views

Some tables may contain sensitive data that shouldn't be seen by all users, or rows that aren't useful for some parts of the user community. So, although you may need certain users to see all the contents of a table, others should see only certain columns or rows, or even a subset of rows and columns. You can accomplish this by creating a view that contains only the elements that should be seen by the selected users and then granting them access to the view rather than to the table. For a view that contains a subset of the columns, you can use the same approach as you would to create a view to hide a dropped column (the example in the preceding section shows the creation of such a view). A view that shows users only a subset of the rows is built by using an appropriate WHERE clause. You can restrict column and row access by building a view with a SELECT clause to identify just the required rows and a WHERE clause to choose the desired rows. The following command shows the creation of such a view. It's based on the EMPLOYEE table from the earlier section "Using Views to Prebuild Queries," but includes only employees in Department 103. Therefore, it doesn't show the department column, nor does it include salary information: CREATE VIEW dept_103 AS SELECT id, last_name, first_name, middle_initial, title, phone, hire_date FROM employee WHERE department = 103 /

Hiding Complicated Queries

Your users and applications may need to execute fairly complicated queries that contain multiple table joins, or subqueries, and combinations of these. If such a query is needed on a regular basis, you might consider creating a view that embodies the query. The user or application can then simply query the view without being concerned with the complexity of the underlying query. This will reduce the possibility of error as well as save time. The following code builds a view that could be used for an online phone directory service based on the EMPLOYEE table (for the name and phone number) and the DEPARTMENT table, which it references (for the department name). Although it's not very complicated in terms of the number of columns and tables involved, it does provide a standard format for the output, using SQL functions and operators: CREATE VIEW phone_list (name, department, phone) AS SELECT UPPER(last_name) || ', ' || INITCAP(first_name) || ' ' || UPPER(middle_initial) || '.', department_name, e.phone FROM employee e, department d WHERE department = d.id / Although the phone listing might be usefully ordered by the NAME field, a view can't contain an ORDER BY clause. A query against the view PHONE_LIST (created in the previous command) to show an alphabetical listing of all employees' phone information would have to include its own ORDER BY clause. The command would be SELECT * FROM phone_list ORDER BY name; The ordering column is the name from the view definition, not from the columns in the base table on which it's based.

Accessing Remote Databases Transparently with Views

To access a table on a remote database, a statement needs to identify the table name plus a database link name

http://www.informit.com/content/0789716534/element_007.shtml (19 of 21) [26.05.2000 16:47:44]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

for Oracle to find the correct remote database and table. (For information on database links, see the Oracle8 Server Distributed Systems manual.) The link name is concatenated to the table name with a commercial "at" (@) symbol. To hide this structure from users and applications, you can create a view that embodies the table and link name. If you needed to reach the EMPLOYEE table in San Francisco from a different database on the network, you could create a database link named SF to point to the San Francisco database and then build a view to hide this link's use. The following shows one version of the command to build the link and then the command to build the view: CREATE DATABASE LINK sf CONNECT TO emp_schema IDENTIFIED BY emp_password USING 'sfdb' / CREATE VIEW employee AS SELECT * FROM employee@sf / Obviously, you can create a view to meet any one of a number of requirements. In some cases, you may need a view to help with a number of issues. There's no reason that the view to access the remote EMPLOYEE, created in the preceding code, couldn't also restrict access to the salary column while renaming the ID column TASK_ASSIGNEE.

Creating and Handling Invalid Views

In some circumstances, you may need to create a view but find that you don't have the permissions on the base table, or that the table you know should exist isn't available. In such cases, you can create an invalid view that will remain unusable until the underlying table(s) is accessible by you. To do this, use the keyword FORCE following the CREATE keyword in your command to define the view. A view can also become invalid at a later time due to changes in the underlying table. Whenever a view is invalid, Oracle will return an error message if you try to execute a statement that refers to the view. After the underlying problem is fixed, the view should work normally again.

Dropping and Modifying Views

Dropping a view simply requires issuing the DROP VIEW command to remove the view definition from the data dictionary. No space will be recovered because a view doesn't own any data. Once in a while, you may need to change the definition of a view. Although this can be done by dropping the view and then re-creating it with the new definition, this might not be a desirable approach. If the view has been granted to users, dropping the view will lose the privileges. Dropping a view that's referenced in triggers or stored procedures causes these objects to be invalidated, requiring a recompile attempt next time they're used, even if you add the new view definition immediately. To preserve the integrity of the users and objects dependent on a view, you can modify the view's definition without having to drop it first. This is done with the OR REPLACE option of the CREATE command. By issuing the CREATE OR REPLACE VIEW... command, you can use a different SELECT clause from that in the existing view without the view disappearing from the data dictionary at any time. You can even modify a view such that a valid view becomes invalid or an invalid view becomes valid, or modify an invalid view to become a different invalid view. If the resulting view would be invalid, you must include the FORCE keyword after the CREATE OR REPLACE clause. Otherwise, the view won't be changed from its previous definition. Don't expect users to be able to use invalid views Just as when you create a view with the FORCE option, any modification that requires the FORCE keyword or that otherwise makes a view invalid renders the view unusable. Nobody can use the view name in a SQL statement successfully until the view is revalidated.

Updating Data Through Views

Although views are used primarily for queries, they can be used for other types of SQL statements. Rows can be inserted and deleted via a view on a single table when there are no set or DISTINCT operators in the view, nor any GROUP BY, CONNECT BY, or START WITH clauses. For row inserts, no columns excluded from the view can be defined as NOT NULL or be part of the PRIMARY KEY constraint (which implies NOT NULL). Updates can also be performed via a view on a single table and, in some cases, through a view on joined tables.

http://www.informit.com/content/0789716534/element_007.shtml (20 of 21) [26.05.2000 16:47:44]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Tables From: Using Oracle8

For a single table, updates are limited by the same restrictions as inserts and deletes. In the case of a view across a join, only one table can be updated in a single statement. Furthermore, there must be a unique index on at least one column in the joined view, and the columns from the table being updated must all be updatable. To see whether a column in a view is updatable, you can query the table DBA_UPDATABLE_COLUMNS (or USER_UPDATABLE_COLUMNS). Understanding Oracle terminology for updatable join views Oracle uses the term key-preserved tables when discussing the update options on views involving table joins. A table is key-preserved in a join view if every key of the table, whether or not it's included in the view's SELECT clause, would still be a valid key following a change to the columns seen in the view. Only key-preserved tables can be updated through the view.

View Consistency
As we've seen, you can create views to restrict the visible rows in the base table. Also, you've learned that you can update a view on a single table. One concern you might have is how these two characteristics work together. Suppose that I use the view DEPT_103, created earlier in the section "Hiding Data with Views," and update it. If I update the employee's title, there shouldn't be a problem. But what if I update one record to change the department number to 242? Now the row doesn't belong to the view and may not be a row I can officially see. You can add a refinement to views that restrict access to certain rows within a table. This refinement prevents users from modifying a row that they can see through the view to contain a value that they aren't allowed to see. This is done by including the key phrase WITH READ ONLY or WITH CHECK OPTION to the view definition. WITH READ ONLY doesn't allow any changes to be made to the base table through the view, so you can't perform an insert or a delete, or complete any updates, on the underlying table. WITH CHECK OPTION, on the other hand, does allow any of these operations as long as the resulting rows are still visible under the view definition. If you want, you can give WITH CHECK OPTION a name by using a CONSTRAINT keyword, just as for other types of constraints (see Chapter 17). The following command shows how you can create a view with a named CHECK OPTION: CREATE VIEW dept_103 AS SELECT id, last_name, first_name, middle_initial, title, phone, hire_date FROM employee WHERE department = 103 WITH CHECK OPTION CONSTRAINT dept_103_dept_ck / The name given to the CHECK OPTION here follows a suggested naming standard developed for constraints. SEE ALSO For more information on naming constraints, < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_007.shtml (21 of 21) [26.05.2000 16:47:44]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Adding Segments for Different Types of Indexes

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

Why Index?
r

The Mechanics of Index Block Splits Sizing an Index Creating an Index Unique Indexes Parallel Operations on Indexes Logging Index Operations Index Tablespaces Index Space-Utilization Parameters Creating Indexes at the Right Time Monitoring Space Usage Rebuilding an Index Dropping an Index Bitmap Index Internals Using Bitmap Indexes Building a Bitmap Index Creating a Reverse-Key Index Rebuilding Reverse-Key Indexes Why Index-Organized Tables Don't Support Additional Indexes Creating an Index-Organized Table Monitoring Index-Organized Tables

Managing a Standard B*Tree Index

r r r

Index Sort Order

r r r r r r r r

Managing Bitmap Indexes

r r r

Managing Reverse-Key Indexes

r r

Managing Index-Organized Tables

r r r

Why Index?
Although the primary reason for adding indexes to tables is to speed data retrieval, you may use indexes for these additional reasons:

http://www.informit.com/content/0789716534/element_008.shtml (1 of 17) [26.05.2000 16:48:03]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

Why indexes are important Imagine looking for a document in a filing cabinet that contains documents in a random order. You might have to look at each and every document before finding what you're looking for. The effort required to find the document will increase as the size of the filing cabinet and the number of documents within it increases. A database without an index is similar to such an unorganized filing cabinet. More than 50 percent of systems reporting a performance problem suffer from lack of an index or from the absence of an optimum index.
q q q q

To enforce uniqueness with a constraint To store data in an index cluster To reduce locking contention on a foreign key constraint To provide an alternate source of data

The first and third items are discussed in more detail in Chapter 17, "Using Constraints to Improve Your Application Performance," which is devoted to integrity constraints. Index clusters, mentioned in the second bullet, are covered in Chapter 18, "Using Indexes, Clusters, Caching, and Sorting Effectively." The fourth item needs some additional comments here. SEE ALSO To learn about using indexes with unique constraints, How to use indexes with foreign key constraints, How to create and manage index clusters, Oracle tries to avoid accessing any more blocks than necessary when executing SQL statements. If a query is written in such a way that an index can be used to identify which rows are needed, the server process finds the required index entries. The process usually uses the rowids-pointers to the file, block, and record where the data is stored-to find the required block and move it into memory, if it's not already there. However, if the columns in the query's SELECT clause are all present in the index entry, the server process simply retrieves the values from the index entry and thus avoids the additional block search in the table itself. This technique, the biggest benefit of which is saving time, can also help out if there's a problem with the base table or the file where it's stored. Queries that can be satisfied from index entries will continue to function, even if the base table is unavailable. Figure 8.1 explains the basic concept of locating data with an index. The user looking for specific information first looks for a keyword in the index. This keyword can be easily located because the index is sorted. The index contains the keyword with the detailed information's address. The desired data is quickly located by using this address information. Figure 8.1 : Indexes provide a quick access path to the data. An Oracle index is a structure, maintained as an independent segment, that contains an ordered set of entries from one or more columns in a table. These ordered entries are stored on a set of blocks known as leaf blocks. To provide fast access to any specific value in these leaf blocks, a structure of pointers is also maintained in the index. These pointers are stored on branch blocks. Each branch block contains pointers for a specific range of indexed values. The pointers themselves may point to a leaf block where the value can be found, or to another branch block that contains a specific subset of the value range. Oracle uses a b*tree index structure, which guarantees that the chain (or number of blocks that must be examined to get from the highest level branch block to the required leaf block) is the same no matter what value is being requested. The number of blocks, or levels, in such a chain defines the height of a b*tree. The larger the height, the greater the number of blocks that have to be examined to reach the leaf block, and consequently, the slower the index. Figure 8.2 shows the logical structure of a b*tree index. Figure 8.2 : A b*tree index consists of a set of ordered leaf blocks with a structure of branch blocks to aid navigation to the leaves. When a leaf block fills up, an empty block is recruited to be a new leaf block; some records from the full block are moved into this new block. This activity is called "splitting a block." The branch block pointing to the original leaf block adds a new entry for the split block. If the branch block doesn't have room for the new entry, it also splits. This, in turn, requires that the branch block pointing to it needs to add a new entry for the split block. The very first branch block, called the "root block," is at the top of the index. If it fills up, it too will split, but in this case the original root block and the split block become the second level in the b*tree. A new root block is created, pointing initially to the two blocks that are now at the next level.

http://www.informit.com/content/0789716534/element_008.shtml (2 of 17) [26.05.2000 16:48:03]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

The Mechanics of Index Block Splits

Most indexes will fall into one of two categories-new entries will be randomly inserted between existing values, or new entries will always have a larger value than any existing value. The former is most often found on columns of character data, such as a last name column. The latter would typically be an index on a numeric primary key column, where a sequence generator or some other tool is used to increment the value of each new entry. In indexes where values are inserted in an apparent random order, very few entries will be stored on the very last leaf block, but are placed on a leaf block somewhere in the sorted sequence of leaf blocks. When one of these blocks fills up and has to be split, Oracle can't make any assumptions about future entries that might be added in the same value range. Therefore, it splits the block in the middle, leaving the first 50 percent of the entries on the current block and moving the second 50 percent of the entries to the new leaf block. This 50-50 split maximizes the available space for new entries that, statistically, are just as likely to fall in the first part of the range as in the second. Splitting the last leaf block The algorithm for splitting the last index leaf block is optimal when the index is on a column that contains ever-increasing values, such as sequence numbers. Only the last entry is moved to the split block, leaving the preceding entries in place. On rare occasions, the index entry on a column containing unordered data, such as last names, causes the last leaf block to split. If the value is currently the highest in the index in such cases, the split will place just this entry on the new leaf block. This can lead to a further block split if other values need to be inserted in the former last leaf block. Although this is less efficient than splitting the block 50-50 the first time it fills, it occurs too infrequently to be a significant performance factor. When a new index entry has a higher value than other current entries, it will be added to the last leaf block in the index. When this block fills up and needs to be split, Oracle may not perform a 50-50 block split. If the new entry is the highest value so far, Oracle simply adds a new leaf block and stores just the new record on it. In indexes where all new entries have higher values, this scheme will provide the maximum possible space on the new leaf block. That is where all new entries will be stored. It also saves the overhead of moving 50 percent of the entries from the current leaf block, which would free up space that would never be used anyway.

Managing a Standard B*Tree Index

The most common type of index is a standard b*tree index. We will take some time examining the various characteristics of this index type and learn how to build and manage one. If you've read through Chapter 5 "Managing Your Database Space," or Chapter 17, you're already introduced to some index issues. Chapter 5 emphasizes the general usefulness of placing indexes in their own tablespaces, away from the tables on which they're built. Chapter 17 briefly discusses the management of indexes required to support certain types of constraints. SEE ALSO For information on tablespace usage for different segment types, To see when to use b*tree indexes, Column order in composite indexes can differ from the table The columns in a composite index may be defined in any order, irrespective of their order in the base table definition. They don't even have to be built on columns adjacent to each other in the underlying table. Before proceeding, you need to be introduced to two terms that appear throughout this chapter. A composite index, also called a concatenated index, is an index that you create on multiple columns within a table.

Sizing an Index
As mentioned in Chapter 5 an index is typically smaller than the table on which it's based. This is because the index generally contains only a subset of the columns from the table and thus requires less storage for each entry than is taken by an entire row. If the index is on many or even all columns, though, it will almost certainly be larger than the table it's indexing. This is because each index entry stores not just the data from the table's columns, but also a rowid, which embodies the physical location of the row in the table. In addition, the index stores one or more branch blocks at every level of the b*tree structure. A table does not need any such blocks; it stores only the data itself.

http://www.informit.com/content/0789716534/element_008.shtml (3 of 17) [26.05.2000 16:48:03]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

Another space issue for an index is the use of empty space. When new rows are being added to a table, the rows can be stored on any available block. As blocks fill, rows are placed on empty blocks and continue to be stored there until they too get full. This can't be done with index entries because they're stored in a specific order. Even if a block is almost empty, Oracle can't store an index entry on it if the entry doesn't belong in the range of values now assigned to that block. Following a block split, an index will have two partially full blocks, neither of which can be used for entries outside either block range. A table, on the other hand, doesn't have to move records around when blocks fill up; it simply adds new records to empty blocks, which it can then continue to fill with other new rows, regardless of their values. As discussed in Chapter 7 "Adding Segments for Tables," Oracle provides detailed descriptions of algorithms that you can use to estimate the total space required by various types of segments. As with computations for table sizes, some of the numbers you have to plug into the formulae are estimates. These include the average lengths of fields stored in indexed columns and estimates of how many rows will have NULLs in the indexed columns, if any. The calculations don't take into account how many block splits might occur while data is being added to the index, so they become less and less reliable for sizing long-term growth. The space required for branch blocks is included for in the calculations, but it's simply based on an expected ratio of branch blocks to leaf blocks. The actual number of branch blocks required depends on the number of distinct values in the index, a factor that isn't included in the space estimate. Sizing an index with sample data I recommend that, if you have good sample data, you should consider building a test index with this data and extrapolate the full index size based on the size of the test index. If you don't have good sample data, it's a somewhat pointless exercise to evaluate the index size by using Oracle's calculations- your input will be a guess. You don't really need to know how big an index will be before you create one, unless you're very short of disk space. In this case, you should ensure that you'll have sufficient space for the entire index. Unlike a table, you rarely need to read an entire index from start to finish, so there's no real requirement to keep its blocks on contiguous disk space. Therefore, you don't need to define large extents for an index; you can afford to create it or let it grow via many small extents. Again, you don't need to have very precise sizing predictions if you plan to use small extents-you won't end up wasting too much space even if the last extent isn't very full, something you can't be sure of if you use large extents. If you want to work on the detailed sizing calculations, you can find Oracle's formulae in Appendix A of the Oracle8 Server Administrator's Guide. Reuse table-sizing scripts You may want to look at the sizing section in Chapter 7and review the scripts that compute sizing requirements for tables based on sample data. These scripts can be modified, if you want to use them, to estimate overall index size. SEE ALSO Get table-sizing details on

Creating an Index
In Chapter 18, "Using Indexes, Clusters, Caching, and Sorting Effectively," you learn what criteria help determine how useful an index would be in optimizing queries or other table access. The other main reasons to use an index were summarized at the start of this chapter. In this section it's assumed you've determined that you need a standard b*tree index on an existing table. You look at the syntax and how to use it to build an effective index. Most indexes you build will use the CREATE INDEX command. In Chapter 17 you can find out about indexes that Oracle builds automatically, if they're needed, to support certain integrity constraints. SEE ALSO For information on constraints that require indexes, The syntax for the CREATE INDEX command to build a standard b*tree index on a table is shown in Listing 8.1. Listing 8.1 Creating an index with the CREATE INDEX command 01: 02: 03: 04: 05: CREATE [UNIQUE] INDEX [index_schema.]index_name ON [table_schema.]table_name ( column_name [ASC][DESC] [,...] ) [parallel_clause] [NO[LOGGING]]

http://www.informit.com/content/0789716534/element_008.shtml (4 of 17) [26.05.2000 16:48:03]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

06: 07: 08: 09:

[TABLESPACE tablespace_name] [NOSORT] [storage_clause] [space_utilization_clause]

Numbering of code lines Line numberings were included in Listing 8.1 and other code listings to make discussion about this code easier to reference. The numbers should not be included with any command-line commands, as part of any Oracle scripts, or within SQL statements. CREATE INDEX...ON on lines 1 and 2 are the required keywords for the command. On line 1, UNIQUE creates an index in which every entry must be different from every other entry; by default, an index is non-unique and allows duplicate entries. index_schema is the name of the owner of the index; by default, it's the user creating the index. Finally, index_name is the name given to the index. On line 2, table_schema is the name of the owner of the table on which the index is being built; by default, it's assumed to be in the schema of the user creating the index. Also on line 2, table_name is the name of the table on which the index is being created. column_name on line 3 is the name of the index's leading column. You can include up to 31 additional columns, as long as the total length of an entry is less than half of the Oracle block size for the database. Also on line 3, ASC and DESC are keywords provided for compatibility with standards; they have no impact on how the index is created. You can use only one of these two keywords per column, but you can apply either one to different columns in a composite index. Finally, [,...] indicates that you can include more than one column in the index, naming them in a comma-separated list. On line 4, parallel_clause is one of

Causes all access to index to be serialized Allows some parallel access Set the number of query slaves to be used in an instance to build index in parallel; only one format can be used per statement Set number of parallel server instances to be used when building index with intern-ode parallel operations; only one format can be used per statement On line 5 of Listing 8.1, LOGGING and NOLOGGING determine whether creation of the index and subsequent activities will be logged (LOGGING, the default) or not logged (NOLOGGING) into the redo logs. The additional activities subject to this setting are direct loads through SQL*Loader and direct-load INSERT commands. TABLESPACE tablespace_name on line 6 identifies the tablespace where the index will be created. By default, it's built in the default tablespace of the user creating the index. On line 7 of Listing 8.1, NOSORT is used to prevent a sort when the rows are stored in the table in ascending order by the index key. The CREATE INDEX command will fail if any row is out of order. By default, Oracle assumes that the rows aren't in order and sorts the indexed data. When to use CREATE INDEX's NOSORT option

http://www.informit.com/content/0789716534/element_008.shtml (5 of 17) [26.05.2000 16:48:03]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

Oracle's tables, like all relational tables, aren't guaranteed to be stored in any specific order. For NOSORT to work when you're creating an index, the table must have been loaded by using a single process with no parallel operations, and with a source of data already sorted in the order of the indexed column(s). The rows can be entered manually, one row at a time with INSERT statements, or with SQL*Loader in conventional or direct mode. The index needs to be created following such a load, before any addition-al DML statements are issued against the table; the row order may not be preserved by such commands. The storage_clause on line 8 is as follows: STORAGE ( [INITIAL integer [K|M]] [NEXT integer [K|M]] [PCTINCREASE integer] [MINEXTENTS integer] [MAXEXTENTS [integer|UNLIMITED]] [FREELISTS integer] [FREELIST GROUPS integer] [BUFFER POOL [KEEP|RECYCLE|DEFAULT]] ) STORAGE is the required keyword. With no STORAGE clause, or for any optional storage components not included in the STORAGE clause, the value will be inherited from the tablespace's default settings. INITIAL integer is the size, in bytes, of the first extent. K and M change bytes to kilobytes or megabytes, respectively. NEXT integer is the size of the second extent. PCTINCREASE integer is the multiplier applied to the size of each subsequent extent following the second extent. MINEXTENTS integer is the number of extents built when the index is created. MAXEXTENTS integer and MAXEXTENTS UNLIMITED are the maximum number of extents allowed for the index, where you must provide a number or the keyword UNLIMITED, but not both. FREELISTS integer is the number of freelists assigned to the index; the default value is 1. FREELIST GROUPS integer is the number of freelist groups assigned to the index; the default value is 1. BUFFER POOL defines the default buffer pool for the index blocks. Only one option is allowed:

q q q q

q q

q q

Order of columns in a composite index If you're building a composite index and aren't sure which columns will be referenced most often, create the index with the columns ordered from the most to the least discriminating. For example, an index on the honorific (Mr., Ms., Dr., and so on), the first initial, and the last name columns: Put the last name first (many different values), the first initial (26 values), and the honorific (a handful of values). KEEP RECYCLE DEFAULT Assigns blocks to the kept buffer pool Assigns blocks to the recycle buffer pool Assigns blocks to neither pool; this is the default if you don't include the BUFFER POOL option

The space_utilization_clause on line 9 is as follows:

Reserves space for new entries on a block (default is 10) Sets number of transaction slots reserved in each block (default is 2)

http://www.informit.com/content/0789716534/element_008.shtml (6 of 17) [26.05.2000 16:48:03]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

Sets maximum number of transaction slots that can be created in a block (default is 255) You don't have to name the columns in a composite index in the same order as they're defined in the table-nor do you, as this implies, have to use adjacent columns. Your best option is to include the most queried column first-a query that provides a value for the leading column of a composite index can use the index to find the required rows, even if the indexed columns aren't referenced by the query. You should include the other columns in descending order of frequency of reference for the same reason-Oracle can use as many of the leading columns of an index as are identified in an SQL statement's WHERE clause. SEE ALSO Find out more about parallel operations, For information about SQL*Loader and its options, including direct and parallel direct loads,

Unique Indexes
You should rarely need to create an index as UNIQUE. A unique constraint, rather than a unique index, should be used to enforce uniqueness between rows. A unique constraint does use an index, and you can create one for this purpose as discussed in Chapter 17, but it doesn't have to be a unique index. A composite unique index will ensure only that the set of values in each entry is distinct from all other values. It will allow the same value to be repeated in a column multiple times, as long as at least one other column has a different value from any existing entry. An entry will be stored in the index if at least one column has a non-NULL value. A NULL value in a column will be treated as potentially containing the same value as another NULL value in that same column. Consequently, an entry containing one or more NULLs, but with all the same values in the non-NULL columns, would be considered in violation of the unique condition. A row with these characteristics therefore couldn't be stored. NULLs aren't considered for uniqueness A row with a NULL in the indexed column won't be recorded in the index, so a unique index won't prevent multiple rows with a NULL in the indexed column from being stored.

Index Sort Order

All indexes are created with the values in ascending order, regardless of the ASC or DESC options settings on each column. Internally, indexes can be scanned in either direction, due to forward and backward pointers between the leaf blocks. ASC and DESC index options The options for ordering indexes in ascending or descending order via the ASC and DESC keywords are included only for compliance with SQL standards. Retrieval from an Oracle index in either direction is equally efficient.

Parallel Operations on Indexes

Unless your index is partitioned, only the creation of the index can be processed in parallel. Subsequent access to the index will be done through a single serial server. Without the parallel clause in the CREATE INDEX command, the creation will also be serial, regardless of the underlying table's definition. The PARALLEL clause's INSTANCES option is of significance only if you're running Oracle Parallel Server. Check availability of parallel server processes Parallel operations on indexes, as well as any other Oracle objects, will occur only if your instance (or instances) is running with a sufficient number of parallel server processes available. This is deter-mined by the PARALLEL_ MIN_SERVERS and PARALLEL_MAX_SERVERS parameter settings in your initialization file. Parallel creation is usually much faster than serial creation, particularly if you have multiple CPUs on your database server and the table is striped across multiple disk drives. For this reason, you may want to build all your indexes with an appropriate degree of parallelism (and number of instances, if applicable). In particular, you should consider creating any indexes needed to enforce constraints with the CREATE INDEX command so that you can include the PARALLEL clause. You don't have the option of defining a parallel operation when an index is built automatically as a result of creating or enabling the constraint. SEE ALSO

http://www.informit.com/content/0789716534/element_008.shtml (7 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

Read about partitioned indexes and their parent segments,

Logging Index Operations

You can speed the processing of your CREATE INDEX commands by not creating redo log entries. After you're done, however, you should try to back up the tablespace where the index is stored-recovery following a media failure can't reconstruct the index correctly without the normally created redo log entries. In addition, you should repeat the backup any time you perform one of the other operations that won't be logged due to the NOLOGGING setting, for exactly the same reason. Of course, in some cases you may not realize that a subsequent operation isn't generating redo entries. Index activities affected by the NOLOGGING option In addition to the creation of the index, the two other activities that aren't logged when the NOLOGGING option is in place are direct loads with SQL*Loader and direct load inserts with the INSERT...SELECT command. To avoid any problems with future unlogged changes to your index, you might want to turn on logging after its creation. You can still create the index without logs-use the ALTER INDEX command after it's built. If you don't include the LOGGING or NOLOGGING option in the CREATE INDEX command, the index will be built in, and future activities will use logging mode for non-partitioned indexes. Partitioned indexes will acquire their logging mode from the parent segment.

Index Tablespaces
By default, indexes are created in your default tablespace just like any other segment. Chapter 5explains why you should consider using different tablespaces for tables and for the indexes on them. Your default tablespace is typically the one where you build your tables, which probably isn't where you want your indexes. It's important to consider where you really want any new index to be created and, if necessary, to include the TABLESPACE clause in your CREATE INDEX command.

Index Space-Utilization Parameters

The space-utilization parameters behave differently for indexes than for tables, which can be confusing: q PCTFREE sets a percentage of each block aside for use by new index entries. Unlike a table, however, this space is reserved only during the initial index creation. After that, the block will fill up as much as possible-even though each new entry will exceed PCTFREE-because index entries must be stored in a specific order. If the initial load filled every block to its capacity, later insertions into the range of values on any block would cause a block split. To reduce the number of such splits, you should reserve space for new entries in indexes where they're likely to occur. Generally, new entries are likely to be needed between other values in alphanumeric data, such as people's names, telephone numbers, or street addresses. Numeric primary keys generally consist of an increasing numeric value, so all new entries will go to the end of the index. Such indexes don't need to reserve free space for new entries into existing blocks. q Because an index block has to contain entries within a certain value range, deletions from the block don't necessarily make the block available for new entries. The space will be reused only if a new entry has a value that allows it to fit between the dropped entry's adjacent entries. No PCTUSED space-utilization parameter is required to control this behavior. q INITRANS has to have a minimum value of 2, not 1 (as in a table). The second transaction slot is required if the block has to be split. This operation occurs during the user's transaction, which is assigned to one transaction slot and is performed by a second transaction initiated by SYS. SYS is assigned to another slot.

Creating Indexes at the Right Time

Unless you have a table that has all its rows physically stored in the order of an index that you need (in which case you can use the NOSORT option), the indexed columns will have to be sorted as part of the index creation. The only way you can guarantee that rows will be in the required order is if you sort them first and load them in that order, after which you prevent any further DML against the table until the index is built. After an index is built, any changes to the underlying table that affect the indexed columns will be automatically included in the index. This adds to the overhead of processing the statements. Unless you need the index to speed the processing or to support concurrent query processing, you should consider dropping it during periods

http://www.informit.com/content/0789716534/element_008.shtml (8 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

of heavy activity against the table. Drop indexes that have frequent block splits Indexes prone to block splitting are good candidates for dropping during periods of heavy DML-either because they're old and have no more free space for new rows on many of their blocks, or because their entries are frequently updated (which results in a DELETE of the old entry and an INSERT of the new, so that ordering is preserved). They can be recreated immediately or, if performance doesn't suffer too much without them, when the underlying table is no longer so busy.

Monitoring Space Usage

As mentioned earlier, the space freed in an index when an entry is deleted isn't available to any new entry added. The "hole" left by the deleted entry is in a specific location in the sort order of the index. For a simple analogy, consider a file cabinet. The drawers correspond to an index leaf block and the folders within are akin to index entries, assuming that they're filed in alphabetical order. If you discard a folder labeled BROWN, it leaves room in the file cabinet between the folders labeled BRONX and BRUIN. If you need to file a new folder for GREEN, you won't place it in the drawer from where the BROWN folder was taken, but in the drawer with the other GR folders-say, between GRAY and GREY. However, a new folder for BROWNING could legitimately occupy the space vacated by the old BROWN folder. An index works in much the same way. Freed space can be reused only if the entry's value allows it to fit into the value range opened by the deleted entry. Index leaf entries aren't dropped only when their corresponding row is dropped from the table, but also when the row is updated, if the update affects the indexed column(s). An update is equivalent to taking a folder from the theoretical file cabinet, changing the label (say, from BROWN to WHITE), and then refiling it. The folder is obviously not going to be of any use filed in its original position; it has to be inserted in the correct position among the other folders beginning with W. As far as the index is concerned, the old entry is removed and a new one is created. The space left behind by the removed entry is subject to reuse in the same way as space left by an actual record deletion. As you can possibly see, an index on a table that undergoes a lot of updates and deletions can end up with lots of free space that may or may not become reusable, depending on the nature of new records being inserted into the table. One type of table and corresponding index will very likely generate lots of free space on the leaf blocks. This is a table with a primary key based on some form of sequence number, either obtained from an Oracle sequence generator or created by the application. Consider an ORDERS table that has an index on the ORDER_NUMBER column and where each new order is given the next highest unused number. Orders are added to the table when they're received and dropped after having been filled and the invoice paid. Over time the leaf blocks with the older orders start emptying out as they're filled and paid. In most cases they will become completely empty, and then they can be recycled for use with new orders. What if some are from delinquent customers who have never paid the bill, or for standing orders that remain in the system for years? The leaf blocks containing such order numbers will have to be maintained, even if there's just one order on them. The free space on these blocks can't be reused by new orders because they will have numbers much higher than the range reserved for these blocks. Consider order numbers 1023 and 3345 on two separate index blocks (on blocks 22 and 28, for example). The blocks between these blocks-23 through 27-may have already been emptied and recycled with higher order numbers. However, the pointers between adjacent leaf blocks that allow an index to be scanned in ascending or descending order will "connect" blocks 22 and 28. Any order outside the range of 1023-3345 therefore can't be placed on either block because it would be out of logical order. It's possible for an index to gradually become burdened by a large number of almost empty blocks. Such an index wastes disk space and results in slow access for values on the sparsely populated blocks. This chapter discusses an index option (reverse-key indexing) that can help you avoid this type of situation. You should regularly evaluate the space usage-and hence, efficiency-of your indexes. Over time you'll learn which indexes are prone to space problems and which aren't, due to their underlying tables being relatively static or the indexed values being random and, hence, able to reuse space. You can use an ANALYZE command's option to see how well or how poorly an index is using its space, particularly with respect to deleted entries. Using the INDEX_STATS view

http://www.informit.com/content/0789716534/element_008.shtml (9 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

INDEX_STATS is a temporary view created by the ANALYZE INDEX...VALIDATE STRUCTURE command. It exists only for the duration of the session that created it and can contain information for only one index at a time. If you execute a second ANALYZE INDEX...VALIDATE STRUCTURE command in your session, the INDEX_STATS view will contain only information about the second index analyzed. Only the session that created it can see the INDEX_STATS view, so another user-or even your userid connected to a different session-won't see the view. When you log out of your session, the view is removed and you'll need to rerun the ANALYZE command to recreate it. The following command will populate the INDEX_STATS view with statistical information about the index: ANALYZE INDEX [schema.]index_name VALIDATE STRUCTURE Of particular interest are the columns LF_ROWS and DEL_LF_ROWS, which show the current number of entry slots in leaf blocks. They also show the total number of entries deleted from leaf blocks, respectively, and LF_ROWS_LEN and DEL_LF_ROWS_LEN, which show the total number of bytes associated with these entries. A rule of thumb is that when the number of, or space used by, deleted entries is greater than 20 percent of total entries, you should consider rebuilding the index to reclaim the space. However, you should also check the PCT_USED column. If this is 80 percent or more-an average amount of space you can expect to see used in a typical index-you may not want to incur the work of rebuilding the index. You should continue to monitor it, however, to ensure that the statistics stay in the preferred ranges. Monitoring the number of keys (leaf entries) versus the number of levels in the b*tree over time is another measure you can apply to an index to see if it's becoming overburdened with deleted entry space. The latter is shown under the HEIGHT column of INDEX_STATS and shouldn't change if the total number of index entries stays the same. If the index height keeps increasing, it indicates that more branch block levels are being added. This behavior is to be expected if more entries are being stored in the leaf blocks. If, on the other hand, the additional branch levels are supporting the same number (or thereabouts) of leaf entries, the structure is becoming top-heavy with branch blocks. This occurs when branch blocks are being maintained for partially emptied leaf blocks. Build a history of index statistics If you want to keep a record of an index's statistics over time, you can issue the command CREATE TABLE table_name AS SELECT * FROM index_stats, where you use a date or sequence number as well as the index name as part of table_name. Remember to do this before you end your session or issue another ANALYZE command. The statistics in INDEX_STATS aren't used by the Oracle optimizers, and the existence of the view in a session won't change the default optimizer behavior. This behavior is different from the statistics collected with the ANALYZE INDEX...COMPUTE STATISTICS or ANALYZE INDEX...ESTIMATE STATISTICS commands. However, if you use these commands, you'll see slightly different values in the DBA_INDEXES than you see in INDEX_STATS. This is because some values in the latter may reflect rows that have been deleted, whereas the values in the former are based only on the current index contents. SEE ALSO To learn more about the use of statistics by Oracle's optimizer,

Rebuilding an Index
You may have a number of reasons to rebuild an index. Here are some of the more common reasons: q To reclaim storage taken by deleted entries q To move the index to a different tablespace q To change the physical storage attributes q To reset space utilization parameters You can use two methods to make these changes. The first is to drop the index and recreate it by using the CREATE INDEX command discussed earlier in this chapter. The second is to use the REBUILD option of the ALTER INDEX command. Each method has its advantages and disadvantages, which Table 8.1 summarizes. Table 8.1 Alternatives for recreating an index Drop and Rebuild: Can rename index Can change between UNIQUE and non-UNIQUE Use REBUILD Option: Can't rename index Can't change between UNIQUE and non-UNIQUE

http://www.informit.com/content/0789716534/element_008.shtml (10 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

Can change between b*tree and bitmap Needs space for only one copy of the index Requires a sort if data exists Index temporarily unavailable for queries Can't use this method if index is used to support a constraint

Can't change between b*tree and bitmap Needs space for duplicate index temporarily Never requires a sort Index remains available for queries Can use this method for an index supporting a constraint

The biggest advantage to dropping and recreating your index is that you don't need space for the original index and the new index to exist at the same time. However, you can't assume that this means the process has no overhead. To build the new version of the index, Oracle will have to perform a sort of the column data in all the existing rows. This will require memory and, for large tables, may even require the use of temporary segments on disk. The sort process will also be time-consuming for a large table, and the index will be unavailable between the time it's dropped and new version is ready. As you may guess, the sort space overhead and time for the work to be done are the biggest disadvantages to this approach. If you elect to use the drop-and-recreate approach to rebuild an index, you need to issue the DROP INDEX command (discussed in the next section), and then use the appropriate CREATE INDEX command (discussed earlier in this chapter). If the index is now being used to enforce a constraint, you can't use this method-Oracle will prevent you from successfully dropping the index. Of course, you can temporarily disable the constraint as long as you're prepared to deal with any changes to the table that may prevent you from reenabling again. The biggest advantages and disadvantages to the rebuild option are exactly the opposite of those for the drop-and-create option. When rebuilding an index, Oracle simply reads the leaf block information, which is already in sorted order, to create the new index. When the index is built, it drops the old copy automatically. Because a sort isn't required, the process is relatively fast. Also, it leaves the original index in place for use by queries that may occur concurrently with the rebuild. The disadvantage is that you must have room in your database for the current and new versions of the index simultaneously. This shouldn't be a problem if you're moving the index to a different tablespace, but may be a deterrent if you need to use the same tablespace. The syntax of the ALTER INDEX...REBUILD command is as follows: ALTER INDEX index_name REBUILD [parallel_clause] [NO[LOGGING]] [TABLESPACE tablespace_name] [NO[REVERSE]] [storage_clause] [space_utilization_clause] The parallel_clause takes the same format as that discussed with the CREATE INDEX command. Here it determines whether the rebuilt operation itself can be done in parallel. If it can, each parallel server process will be responsible for retrieving a subset of the current entries and building the new leaf blocks for them. This may cause more extents to be used than a serial rebuild because each slave process will create and use its own extents. Some of these extents may be trimmed back at the end of the operation to remove any unused blocks. The REVERSE/NOREVERSE option determines whether the replacement index is (REVERSE) or isn't (NOREVERSE) a reverse-key index. You read more about reverse-key indexes later in this chapter; for now, simply note that this option allows you to build the replacement either way, regardless of how the current index is structured. The TABLESPACE option can be used if you want to move the index to a different tablespace. If you don't include this option, the index will be rebuilt in the same tablespace as the original index, not in your default tablespace. Back up your work after executing commands without redo entries If you select the NOLOGGING option, you may want to back up the tablespace in which the new index resides as soon as possible. This option will preclude the creation of redo log entries for the replacement index build just as it does when used with the CREATE INDEX command, with the same consequences discussed. The other clauses in the command all work exactly as they do for the CREATE INDEX command. You can refer to the earlier section, "Creating an Index," where this command and these options are explained in detail.

http://www.informit.com/content/0789716534/element_008.shtml (11 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

Dropping an Index
Unless your index was created for you when a constraint was enabled, you should be able to drop it at any time. You may decide to drop an index for any number of reasons, including the need to recreate it (as discussed in the previous section). You may also want to drop an index because it's no longer being used by the statements accessing the table, or because you're going to perform a major data load (or other intensive DML activity) and don't want to incur the overhead of concurrently maintaining the index. Parallel direct data loads, in particular, can't maintain any indexes on a table, so you should always drop them before using this loading technique. The syntax for dropping an index is very straightforward: DROP INDEX [schema.]index_name You need to include the schema name only if the index doesn't belong to you. If you need to drop an index that's supporting a constraint, you must first disable or drop the constraint. You can find details on these steps in Chapter 17. If the index had been created automatically as part of the constraint definition and enabling, it will also be dropped automatically.

Managing Bitmap Indexes

A bitmap index is a special type of index particularly suited to large tables and to columns with a small number of distinct values. Their structure also makes them very efficient when two or more conditions have to be met by a query and the columns involved each have a bitmap index. This structure, however, makes them less efficient than a regular b*tree index when indexed columns are updated, or when rows are added or deleted from the table. Bitmap indexes behave like other indexes when created After a bitmap index is created, it works behind the scenes just like any other index. The users don't have to do anything different to get a bitmap index to be used by a query than they do for a standard index. In addition, bitmap indexes are maintained transparently as records in the underlying table are inserted, updated, or deleted. SEE ALSO For when to use bitmap indexes,

Bitmap Index Internals

A bitmap index is actually stored inside a regular b*tree index structure. The entries in the leaf blocks consist of a value, the rowid of the first row represented in the bitmap, the bitmap itself, and some entry-management bytes. These leaf block entries are stored in order by value, just as in a regular b*tree index, and the index can also contain branch blocks to speed access to a required leaf block value. Even more bitmap index internals Before storing them, Oracle applies a compression algorithm to the bitmaps; this reduces the space required to store consecutive bits with the same value. Although this makes the index more compact (and, hence, reduces the number of blocks that have to read into memory when it's being scanned), it slows the process of changing a bit's value. To help reduce the overhead of managing com-pressed bitmaps, each bitmap actually stores a bit for every potential row in its range rather than just for stored rows. In this way, only the bitmap containing the value for the indexed column( s) of a newly inserted row will need to have its bitmap changed, flipping the bit for the corresponding row from 0 to 1. The other bitmap values stay unchanged. The value portion of the bitmap entries correspond to the values from the table's indexed column. A small table will have one entry for every distinct value in the column, including an entry for NULLs. The bitmap for each entry will contain a bit for every row in the table. Each bit will be set according to the value in the corresponding row-a 1 if the column's value matches the index value, a 0 if it doesn't. This means that there will be only one index entry with a 1-bit for any given row in the index. Figure 8.3 shows a logical view of a bitmap index on a column with four different values. Figure 8.3 : A bitmap index consists of index entries (values) and strings of bits (bitmaps) stored in a b*tree structure. For larger tables, where bitmap indexes are most effective, a single bitmap can't store a bit for every row in the

http://www.informit.com/content/0789716534/element_008.shtml (12 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

table. This is because each index entry-including the value, the bitmap, and the overhead bytes-must fit into less than half a block, a restriction placed on every Oracle index. To cover a large table, multiple index entries are created for each distinct value. Each entry contains a bitmap that represents a disjoint subset of the rows in the table. These bitmaps work the same way as described earlier-a rowid in each entry points to the first row covered by the entry's bitmap, and each bitmap uses a 1-bit to indicate that the value is in the corresponding row and a 0-bit for all other rows. Although bitmap indexes are typically built on a single column, they can cover multiple columns. Composite bitmap indexes work exactly as previously described, except that each entry or set of equal-valued entries corresponds to distinct values in the combination of columns.

Using Bitmap Indexes

Bitmap indexes were designed to support the type of tables that are becoming common in data warehouse databases-very large, relatively static, and not normalized. They can support large tables easily because they're relatively compact, even on tables with millions of rows. They're best for static tables because they use compression algorithms that require heavy maintenance when indexed values are added, dropped, or updated. The same values will tend to occur frequently in tables that are not normalized, which wouldn't make their columns good candidates for a standard index. Bitmap indexes are even efficient when two or more values are requested from a column because the bitmaps from each value can be combined with a Boolean OR. They're very efficient at finding rows based on multiple column queries, such as the following: WHERE region = 'West' AND week = 43 AND salesman = 202 AND product = 1497 AND color = 'RED' With a bitmap index on each column named in this example, the required rows can be found by performing a Boolean AND on each bitmap. The resulting bitmap will have a 1-bit only in the positions corresponding to rows having all the required values. Boolean ANDs and ORs are some of the most efficient operations a computer can perform. Use a large sample size when estimating bitmap index storage requirements A single bitmap can store only a finite number of bits, representing a fixed number of rows. In a small table, every row can be covered by one bitmap for each distinct value in the indexed column. At some point, the insertion of a new row will require a second bitmap to be built for each value, doubling the number of blocks needed to store the bitmap entries. An insertion will cause the third bitmap to later be built for each value, increasing the number of leaf blocks required by another 50 percent. Until your index is large enough so that the addition of a new set of bitmaps per value doesn't increase its size significantly, you can't extrapolate its final size accurately.

Building a Bitmap Index

To create a bitmap index, you should use the CREATE INDEX command discussed earlier, except that you have to include the keyword BITMAP where the optional UNIQUE keyword appears. All the other options shown in Listing 8.1 for the CREATE INDEX command can be used for a bitmap index. You can also rebuild a bitmap index by using the ALTER INDEX...REBUILD command. The only option shown in the previous section that you can't use for a bitmap index is REVERSE. Of course, you can always drop and rebuild a bitmap index by using the standard DROP INDEX command. One aspect of bitmap indexes that's difficult to manage is planning how big they will be, using a formula such as the one published by Oracle for regular b*tree indexes. The amount of data compression isn't going to be obvious from an examination of sample data. Your best way to estimate the size of the final index is to extrapolate it from an index built on a sample table.

Managing Reverse-Key Indexes

A reverse-key index is one in which the value portions of the entries in the leaf blocks are saved in reverse bit order. The entries themselves are stored in the order of the resulting bit patterns rather than in the order of the original values. The branch blocks, of course, also reflect the stored values-that is, the reversed bit values rather than the values as found in the tables. The result of this bit reversal is to make somewhat random the locations of entries throughout the index. Values that normally would be stored sequentially are likely to be scattered

http://www.informit.com/content/0789716534/element_008.shtml (13 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

throughout the index on different blocks. You use a reverse-key index to help avoid the problem of having an index develop lots of empty spaces because entries were dropped and no new records could reuse the space. This is likely to happen in a normally organized index when new records have increasingly higher values, such as sequence numbers, but not all the older entries are removed from other blocks. A good example of this type of index is one on an employee ID column. In a growing company, more new employees are added than existing employees leave. The remaining employees have index entries for their ID numbers on the older leaf blocks, which are probably partially empty as a result of some earlier employees leaving. However, the new employees' ID numbers will be too high in value to be stored within the value ranges of these blocks. Over time, the index may become inefficient due to increasing amounts of unusable free space. As discussed earlier, you would have to drop and rebuild the index to compress the earlier values into full blocks. Rather than have to compress the index regularly, you could build it as a reverse-key index, which should spread high and low values around the used blocks. When entries are removed, new ones stand a reasonable chance of fitting into the ranges opened by the deletions. SEE ALSO To learn when to use reverse-key indexes,

Creating a Reverse-Key Index

Use the CREATE INDEX command shown in Listing 8.1 to create a reverse-key index, but include the keyword REVERSE where the optional NOSORT keyword appears. All the other options shown in Listing 8.1 for the CREATE INDEX command can be used for a reverse-key index. To convert an existing index into a reverse-key index, use the ALTER INDEX...REBUILD command as described earlier. Use the keyword REVERSE as shown in that section. Of course, you can always drop a regular index and rebuild it as a reverse-key index, using the standard DROP INDEX command and the CREATE INDEX command as just described.

Rebuilding Reverse-Key Indexes

To rebuild a reverse-key index, you can drop the original index and create it again. If you decide to use the ALTER INDEX...REBUILD command, you can make any other changes you need, such as changing the tablespace or storage values. Changing the type of index during a rebuild A reverse-key index will be rebuilt as a reverse-key index whether or not you use the optional keyword REVERSE. However, if you use the NOREVERSE keyword, your rebuilt index will no longer be a reverse-key index, but will instead be a standard index.

Managing Index-Organized Tables

An index-organized table is a hybrid structure consisting of b*tree index and regular table elements. The b*tree components are the use of ordered entries on leaf blocks, and branch blocks that provide fast access to the required leaf blocks. The table-like features include the use of standard SQL commands for table management, such as creation and alteration, and the use of standard DML and SELECT commands to manipulate rows. (See Chapter 7for information on table creation and management.) The complete rows-not just the indexed column data-are stored in the b*tree block structure. Index-organized tables provide the fast data access that's normally achieved only by creating an index on a table but don't require the storage space, or the management, of a separate index. You should consider using an index-organized table for lookup tables, which are almost exclusively used to find a value, or set of values, based on a key. For example, a table containing city names you find by entering the postal zip code are an ideal candidate for an index-organized table. Indexed-organized tables do have some restrictions. Unlike a regular table, you can't create additional indexes; therefore it's unlikely that you would want to use this structure for tables with columns with different columns that you typically access by specific value. Remember that UPDATE and DELETE commands on specific rows, not just queries, can benefit from indexes on the columns referenced in the WHERE clause. The ordering of an indexed-organized table is based on its primary key, so the table must have a primary key defined. This, of course, is recommended for all relational tables, but isn't a requirement and makes a second restriction (if you consider it such) on index-organized tables.
http://www.informit.com/content/0789716534/element_008.shtml (14 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

Finally, as with a regular index, the space vacated by a dropped row can be reused only if a subsequently inserted row has the same value, or a value that lies between the adjacent column values. This can result in a table with more free space than you would expect to see in a regular table where the PCTUSED setting controls the reuse of released space. SEE ALSO To see when to use index-organized tables,

Why Index-Organized Tables Don't Support Additional Indexes

An index uses a rowid to point to a specific row in a table. The rowid identifies where the row is physically stored-the block number within that file and, to ensure that each rowid is unique, the row number on that block. It's this last piece that makes rowids unusable for entries in index-organized tables. The row number for a regular table's row is actually the number of the row's entry in the block's row directory. The row directory is simply an ordered list of byte addresses, entry one containing the byte where row one starts, entry two containing the entry where row two starts, and so on. If the row has to be moved to another location in the block due to an UPDATE that extends its length, for example, Oracle simply moves the row and updates its directory entry to reflect its new starting byte address. In an index-organized table, the rows have to be stored in sorted order, just as the entries in an index are stored. If a new row is inserted into the table with a primary key value that lies between two existing values, the row must be logically stored between the two rows containing these values. Suppose that the table contains the values "Baltimore" and "Boston," and that they are in row positions 10 and 11 on a particular block-that is, the tenth entry in block directory contains the starting byte address for the row containing "Baltimore" and its eleventh entry points to the "Boston" row. Index-organized tables can't support unique constraints A unique constraint is enforced by Oracle with an index built on the constrained columns. Due to the limitation that index-organized tables can't have additional indexes, you can't add a unique constraint to such a table. If we add a row with "Biloxi" as its primary key, it now needs to use the 11th slot in the block row directory for its byte address; the entry for "Boston," and any other rows following it, have to be moved up to the next slot. Were the rows being treated as regular table entries, this action would effectively change the rowid addresses for the "Boston" row and the rows following it on the block. Furthermore, if we had been able to build indexes on this table, Oracle would have to find every index entry that pointed to these rows and update the rowid values, even if the rows themselves didn't move by even one byte from their original locations. You can help control space usage within an index-organized table by splitting each row into two pieces, the first of which acts more like a regular index entry and the second of which behaves more like a table row. Oracle will build distinct segments to store the index-organized table when you do this. The first segment will hold the portion of each row that constitutes the indexed, or ordered, values; the second segment will hold the remainder of each row. This approach reduces the amount of wasted space that can occur when a row is dropped or the primary key value is changed, causing the row to be relocated. Only the row piece in the first of the two segments, the ordered segment, leaves space that has to be reused by an equivalent value. The contents in the second segment are treated like the rows in any other table-space can be reused by any other row piece that fits on the block, and an UPDATE doesn't necessarily cause the row to migrate to a new block. To build your index-organized table in two such segments, you begin by determining how much of an average row you want to keep in the index portion and how much in the table portion. Generally, you want the primary-key columns in the former and the balance of the row in the latter. You then need to compute the percentage of the block that the index portion consumes for an average row. When that's done, you can build the index-organized table and identify this percentage with storage information for the table's two parts. You can even identify a column that determines where the break between the index portion and the row portion is made; this will ensure that the entire primary key is kept together, even if its length is less than the specified percentage.

Creating an Index-Organized Table

To create an index-organized table, you use the CREATE TABLE command (discussed in detail in Chapter 7 with some additional keywords and certain restrictions. Listing 8.2 shows the syntax for this command. Listing 8.2 The CREATE TABLE command for an index-organized table

http://www.informit.com/content/0789716534/element_008.shtml (15 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

01: 02: 03: 04: 05: 06: 07: 08: 09: 10: 11: 12: 13: 14: 15:

CREATE TABLE [table_schema.]table_name ([column_description [,...], [CONSTRAINT constraint_name] PRIMARY KEY (column_n|ame [,...])) ORGANIZATION INDEX [TABLESPACE tablespace_name] [storage_clause] [space_utilization_clause] [[PCTTHRESHOLD integer] [INCLUDING column_name] OVERFLOW [TABLESPACE tablespace_name] [storage_clause] [space_utilization_clause] [PCTUSED integer]]

On line 2, column_description includes the column name, the column datatype, an optional size, and an optional constraint. Restrictions on constraints in index-organized tables Because an index-organized table can't have additional indexes, you can't use a unique constraint on any column or column combination. You can include any other type of constraint, however. SEE ALSO To learn about the column-definition options used when building a table, The syntax to add a constraint to a column definition, PRIMARY KEY on line 3 can be a named or an unnamed constraint, defined as a column or a table constraint on a single column or composite key. SEE ALSO For details on primary key constraints, Line 4's ORGANIZATION INDEX are required keywords. On line 8 PCTTHRESHOLD integer sets the percentage of space in a block that any row can consume; the balance of a row that exceeds this threshold is stored in an overflow area. The default value is 50. Interaction of PCTTHRESHOLD and OVERFLOW clauses If you include a value for the PCTTHRESHOLD option and don't include an OVERFLOW clause, any row that exceeds the percentage of block space defined by PCTTHRESHOLD will be rejected. INCLUDING column_name on line 9 names the first column that will be placed in the overflow area, if needed; the column must be the last column defined in the primary key or a non-primary key column. OVERFLOW on line 10 is a required keyword that introduces the definition of the overflow segment on lines 11-through 14: q TABLESPACE tablespace_name, storage_clause, and space_utilization_clause all provide the same options as defined earlier in Listing 8.1's CREATE INDEX syntax. These clauses can contain different values for the primary storage segment (the ordered, index portion of the rows) and the overflow storage segment (where the remaining columns of the rows are stored). q PCTUSED defines a threshold of space used in a block below, which block is returned to the free list.

Monitoring Index-Organized Tables

The data dictionary stores information about index-organized tables in a number of different tables that you can see through the DBA_/ALL_/USER_TABLES and the DBA_/ALL_/USER_INDEXES views. The *_TABLES views contain the basic definition of the table under the name you provide in the CREATE TABLE command. These views also contain information about the overflow segment, which, if present, contains values for the extent sizes and related storage information. The overflow segment also holds the statistics generated from the ANALYZE TABLE command, whether you name the base table or the overflow table name in the command. The *_INDEXES views contain most of the segment's description, including its storage and space utilization
http://www.informit.com/content/0789716534/element_008.shtml (16 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Adding Segments for Different Types of Indexes From: Using Oracle8

values, as well as statistics generated with the ANALYZE command. These views also show the name given to the index in which the rows are actually stored. If you don't specify an overflow option, these views will contain the only detailed information about the physical storage of the segment. Identify the index characteristics' names and, possibly, its overflow segment to find out all the information about an index-organized table in the data dictionary. You can use the following query to find values in the data dictionary columns specific to the objects comprising index-organized tables: Finding the extents belonging to an index-organized table If you query the DBA_EXTENTS data dictionary view to find the segments belonging to an index-organized table, you won't find them under the segment name you gave to the table. Instead, you find them listed under the Oracle-sup-plied segment names for the index segment and, if there is one, the overflow table segment. To relate these extents to a specific index-organized table, therefore, you have to use a query, like the one provided, to find the names of its extents. SELECT t.owner AS "Owner", t.table_name AS "Table Name", i.index_name AS "Index Name", i.tablespace_name AS "Index TS", o.table_name AS "Overflow Segment", o.tablespace_name AS "Overflow TS", i.pct_threshold AS "Overflow Pct", i.include_column AS "Include Col" FROM dba_tables t, dba_indexes i, dba_tables o WHERE i.pct_threshold IS NOT NULL AND t.table_name = i.table_name AND o.iot_name(+) = t.table_name; You can, of course, restrict this query by naming a specific table or a specific owner (for example, adding the clause AND t.table_name = 'table_name' or AND t.owner = 'owner_name', respectively). Of course, after you use this query to identify the names Oracle has given to the indexes and, when appropriate, to the overflow objects, you can use their names to query the *_TABLES and *_INDEXES views to see other information about them. To find out how well a specific indexed-organized table is behaving, you can use the ANALYZE INDEX...VALIDATE STRUCTURE on the index built for the table. You should look for the same criteria discussed earlier regarding determining whether a regular index needs reorganizing. However, if you do find problems with an index for an index-organized table, you can't use the ALTER INDEX...REBUILD command to recreate it. If you determine that it does need reorganizing, you may have to drop and recreate the entire index-organized table. However, you can use the ALTER INDEX command on the index itself or the ALTER TABLE command on the base table name in order to make some changes to the index that you could otherwise make to any other primary-key index with the ALTER INDEX command. You can also monitor the block usage in an index-organized table's overflow segment. To do this, use the ANALYZE TABLE command to compute or estimate statistics on the overflow segment. As mentioned earlier, you can name the base table name or the overflow segment name in the command. If you determine that there are storage problems, you may need to drop and recreate the whole index-organized table. Unfortunately, you aren't allowed to make changes to the overflow segment directly with an ALTER TABLE command in the current release. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_008.shtml (17 of 17) [26.05.2000 16:48:04]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Creating and Managing User Accounts

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q

The Key Purposes of a User Account Creating User Accounts

r r r r

User Authentication Creating a User Account Allowing Quotas on Different Tablespaces Using the CREATE SCHEMA Command Using the ALTER USER Command Understanding the Results of Changing Quota Using the DROP USER Command

Modifying User Accounts

r r r

The Key Purposes of a User Account

User accounts in Oracle8 generally provide three distinct and important purposes in the operation of a database: q They provide security so that you can control who has access to what portions of your database and what operations can be performed. q Even if someone is allowed access to data, you may want to audit his or her activities as part of an overall security plan or to help diagnose performance problems by identifying who is holding locks, using system resources, or simply logged into the database. q An Oracle8 account includes a schema wherein all segments (tables, indexes, stored procedures, and so on) owned by the user are stored. The schema name matches the user's userid. User security is one of the more important functions that you perform. Security in Oracle8 is important not only in keeping prying eyes from your sensitive data, but also in helping keep a developer or normal user from interfering with a production database's operations. Oracle8's security allows you to impose a very granular security profile on every user in the system. This profile allows control over which columns, tables, and rows users can select, update, or insert into, as well as whether users can create segments and where they may be created. Although security is probably one of the more thankless jobs, it's absolutely essential that you understand what types of security protection are available. Every security permission has some risks and benefits associated with it. It's important that the significance of each permission be understood so you can implement the safest and most flexible security policy possible. It's true that nobody will thank you for expiring his or her password every 90 days, but if someone breaks into the corporate data warehouse, you'll probably be answering some even tougher questions from your organization's upper management. Oracle8 improves user management

http://www.informit.com/content/0789716534/element_009.shtml (1 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8

User accounts through Oracle7 had been one of Oracle's traditional weaknesses-until Oracle8. Oracle8 truly has an excellent system in place for managing user accounts (and security in general) throughout your enterprise. It's vastly superior to anything Oracle7 DBAs had to work with.

Creating User Accounts

An Oracle8 user account allows a user to be identified by the database and, optionally, to store segments such as tables and indexes. What users do with their individual user accounts is controlled by permissions that you grant. A user account by itself doesn't allow anyone to even log on to the database. No implicit permissions are given when a user account is initially created. Oracle, like most RDBMS systems, uses privileges granted to a user account to determine whether a user is allowed to perform a certain function on a particular segment (or in the system at all). A user account without any privileges granted can't do anything at all. Granting users privilege is covered in depth in Chapter 10, "Controlling User Access with Privileges."

User Authentication
The only means of identifying users to Oracle is their user identification, known as a userid in Oracle8. Anyone could deliberately or accidentally claim to be someone he isn't. Without some means of authenticating who a user is, you could potentially allow someone access to information; that access could be damaging to the company, employees, or worse. From the earliest of computer days, the password has been a universally accepted means of authenticating who a user is. Central to password authentication is the belief that only a person claiming to be userid GASPERT would know what GASPERT's password is. By making the user produce this secret, you believe that if a user knows the password, he is who he says he is. The popularity of password authentication is also one of its greatest drawbacks. Most users despise having to keep a different password to each system they use during the day. Users are usually urged to use different passwords on different systems, which leads them to write down passwords or use a single password on all systems anyway. In either case, the password can no longer be considered a secret. You have no way of knowing if someone else had access to the password either in another system or to the piece of paper on which the password was written. Without a doubt, it's at least an irritant to have to log on to multiple systems separately each time you want to use them. Clearly, it would be most convenient if you could just log on to one system (such as the operating system) and then have all subsystems (such as a database) base their authentication on the first logon. Fortunately, Oracle8 provides this exact capability. Oracle8 allows external authentication, which basically links a particular Oracle8 user account to a particular operating system account. External authentication's drawback The only major drawback to external authentication is that you have to trust the operating system to authenticate users properly. Don't take this issue lightly. PC environments in particular are notorious for their lack of credible security. By using external authentication, your database's level of security is no greater than your operating system's. If you can trust your operating system to provide adequate security for your database, external authentication offers an attractive and viable alternative to saddling your users with yet another password. In the modern client/server world, users often aren't actually using a tool such as SQL*Plus on the same machine that the Oracle8 server runs on. External authentication then adds another question to ponder: Can we trust the remote operating system to properly authenticate a user? Often, external authentication is practical only if you can trust the database server's OS and the client machine's OS's authentication methods. PCs, again, present a particular problem. Remote authentication and Microsoft Windows Windows 3.x and, arguably, Windows 95 don't offer nearly the level of security that Windows NT or a UNIX workstation offer. If your clients will be using Windows 3.x or Windows 95, you probably should avoid using external authentication. If you want to use operating-system authentication, you need to be aware of two Oracle8 parameters and their customary settings:

http://www.informit.com/content/0789716534/element_009.shtml (2 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8

os_authent_prefix = OPS$ remote_os_authent = true In Oracle8, externally authenticated userids are typically prefixed with OPS$ to distinguish them from users who must log in with the traditional userid/password combination. os_authent_prefix allows you to customize the prefix of externally authenticated userids. If you plan to allow a remote system to authenticate a user for Oracle8, you must set the remote_os_authent parameter in your INIT.ORA file to TRUE. Oracle8 introduced a third option for authentication, known as enterprise authentication. With the Oracle8 Security Service (OSS), this system creates a global central location for authenticating user logins. If your environment consists of many Oracle8 databases, this may be appealing to you. Although you still must create a user account in each individual database a user is allowed to access, the authentication and password is stored in a central area. This allows users to have a single logon to all databases. If users change their passwords, the change affects all database systems that use a common OSS. The installation and configuration of OSS is beyond the scope of this book. The discussions and examples henceforth assume that you aren't using OSS in your environment.

Creating a User Account

Before creating a new user on your database, you must gather some information first. You need to know the following: q What will the new user's userid be? q Will the user be externally authenticated? If not, you must choose an initial password for the user. q If the user will be creating segments, what tablespace will he or she be placed in by default? q When temporary space is needed for operations such as sorting, which tablespace will be used? In many systems, a tablespace named TEMP will be dedicated for this purpose. q How much space, or quota, will the user be allowed to use on tablespaces in the system? q If the user will be assigned a security profile, what profile will be used? To create a user in Oracle8, use the CREATE USER command with the following format: CREATE USER userid IDENTIFIED BY password | EXTERNALLY [DEFAULT TABLESPACE tablespace] [TEMPORARY TABLESPACE tablespace] [QUOTA value | UNLIMITED] ON tablespace] ... [PROFILE profile] [PASSWORD EXPIRE] [ACCOUNT <LOCK | UNLOCK>] Where to use the CREATE USER command This discussion of creating user accounts uses SQL*Plus and assumes that you're logged in as a user with the necessary privileges to create another user. This typically means that you're logging in as the SYSTEM user or have been granted the DBA role.
q

CREATE USER userid This specifies the userid of the new user you're creating. Using any existing username standards in place at your site is advisable to minimize confusion in your user community. If this user will be externally authenticated, be sure to prefix the userid with the value of the os_authent_prefix parameter (typically, OPS$). BY password | EXTERNALLY If you aren't using external authentication, you must provide an initial password for the new user. This password should begin with an alphabetical character (a-z) and consist of alphabetical or numeric characters. Be aware that this password will be visible onscreen. You should avoid creating users at a workstation where others could easily see the password(s) you're entering. If you're using external authentication, you must enter EXTERNALLY after the IDENTIFIED clause.

Who's looking over your shoulder?

http://www.informit.com/content/0789716534/element_009.shtml (3 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8

More than one company's security has been breached by someone merely observing a user account being created. After creating a new account, you should always exit SQL*Plus and clear the screen. Remember that SQL*Plus remembers the last command entered. If you create a new user account, clear the screen, and walk off, someone could enter a quick SQL*Plus command and see exactly what password you used to create the new account. Many companies routinely create new accounts with a well-known password, allowing almost anyone access to your system. This is a poor policy from a security standpoint. Always make new user account passwords unique.
q

DEFAULT TABLESPACE tablespace If this new user account will be allowed to create segments, be sure to set a default tablespace for where these segments will go. This won't restrict the user from creating segments in other tablespaces; it just provides a default location for storage if the user doesn't specify otherwise. If you don't specify a default tablespace, the default is to place user segments in the SYSTEM tablespace. Never allow users to create segments in the SYSTEM tablespace or use it for temporary sort storage. They could bring your database operations to a complete standstill if the SYSTEM tablespace fills up!

Keep the SYSTEM tablespace clean

TEMPORARY TABLESPACE tablespace In the course of sorting functions, Oracle8 will require some scratch space in which it can create temporary segments until a transaction completes. You should dedicate at least one tablespace to this function. Typically, this tablespace will be named TEMP. If you don't specify a location for temporary segments, Oracle8 defaults to the SYSTEM tablespace. QUOTA value | UNLIMITED ON tablespace Oracle8's quota system allows you to limit the amount of space a user's segments may use on any given tablespace. By duplicating the QUOTA clause as many times as necessary within the CREATE USER command, you can set up quotas on all tablespaces in which the user will be authorized to store segments. You should generally assign an appropriate quota on the user's default tablespace. The value parameter is an integer followed by K or M (no space) to denote kilobytes or megabytes. If you don't use K or M, Oracle8 will interpret the integer you provide as bytes, which probably isn't what you intended Oracle8 to interpret. For instance, 40M, 40K, 40 will be interpreted as 40MB, 40KB, and 40 bytes, respectively. PROFILE profile Oracle8 allows you to create standard security profiles that can be assigned to users. This greatly simplifies and standardizes the security policies to be enforced. You can specify a security profile by adding PROFILE followed by the profile name in the CREATE USER command. Oracle8 provides many new security features that can enhance the security of your Oracle8 database. SEE ALSO For a more in-depth discussion of profiles, PASSWORD EXPIRE Oracle8 allows for the initial password to expire immediately. To do this, add PASSWORD EXPIRE to the end of the CREATE USER command. This forces the user to change the password when he or she first logs in to Oracle8. ACCOUNT LOCK | UNLOCK In some environments, it's desirable to create accounts for new employees but not allow access (by LOCKing them) until they're actually ready to use them (at which time they will be UNLOCKed). By adding ACCOUNT LOCK to the end of your CREATE USER command, the user account will be created but the user can't log in until you explicitly unlock the account.

Basic Example of Creating a User The following is a basic example of creating a user in Oracle8. Before entering this command, log on to SQL*Plus as a user with DBA privileges (such as the SYSTEM user).

http://www.informit.com/content/0789716534/element_009.shtml (4 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8

Used to create user RWILSON Sets password to RH0WIL2 Indicates that USERS table-space will store RWILSON's segments by default Set to store RWILSON's temporary sort segments in the TEMP tablespace Set to allow 100MB in USERS tablespace Set to allow no storage at all on SYSTEM tablespace Forces RWILSON to change password on first logon attempt Prevents user from logging on until you unlock account After you enter this command, the new user account is created. If the user tried to log in right now, Oracle8 would stop her because she would lack the CREATE SESSION privilege. Chapter 10 discusses the privileges needed by new users to connect and work in the database. Example of Creating a User with External Authentication and a PROFILE The following example shows how to create a user who will be externally authenticated and has a security profile assigned to him:

Used to create OPS$BRONC (OPS$ is required because this account is externally authenticated.) Indicates that new account will be externally authenticated Set to use security settings in STD_USR_PROFILE profile for new account

http://www.informit.com/content/0789716534/element_009.shtml (5 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8

Allowing Quotas on Different Tablespaces

Oracle8 allows you to control the amount of space users can consume in tablespaces. By default, there is no quota on the amount of space a user's segments may use in the database. Typically, you'll want to restrict users' storage to prevent someone from consuming all available space. Furthermore, you'll probably have at least one or two special-purpose tablespaces that you'll want to reserve for a specific use. This is particularly true of the SYSTEM tablespace, which should never contain user segments. The QUOTA parameter, while straightforward, has the potential of greatly affecting the success and efficiency of your database. Too loosely set parameters will usually result in bloated, duplicated segments that have been poorly sized, resulting in wasted disk space, or worse-not enough free space for everyone to continue working. If QUOTA settings are set too tightly, users may waste time with unnecessary workarounds to conserve space. Coordinating your quota policies No matter what policy you adopt, it's essential that you coordinate it with your users during the initial policy-making and in regular follow-up discussions. Remember that your job as a DBA is to provide a service to your users. The purpose of quotas is to ensure equal quality system performance for all. Setting quotas on a user applies only to segments owned by that user. If the user is allowed to create segments in other users' schemas or insert rows into another user's table, only the segment's owner's quota will be applied. The QUOTA parameter is normally used to provide a fixed limit, such as the following: QUOTA 300M ON USERS There are two special cases, however: no quota limit at all and not allowing any space at all. To remove quota restriction in a particular tablespace, you must substitute UNLIMITED for the numeric limit: QUOTA UNLIMITED ON USERS This allows a user to use as much space as he or she asks for (as long as there's enough free space to supply the request, of course). If you don't specify a quota limit on a tablespace, Oracle8 will default to an UNLIMITED quota. To prevent a user from creating segments on a particular tablespace, set the size to 0 as shown in this example: QUOTA 0 ON USERS Because a segment must be larger than 0, it's impossible for a user to create a segment that will satisfy a 0 quota limit. As explained earlier, you can duplicate Oracle8's QUOTA parameter as many times as necessary within a CREATE USER command. This is useful to easily set quota on many different tablespaces at once. You'll often want to establish quotas on all non-rollback and temporary tablespaces; this is particularly true in production environments. Suppose that a database has the tablespaces SYSTEM, TEMP, RBS, PROD, DEV, and TOOLS. The PROD tablespace is intended to hold production segments, whereas DEV is used to hold developer's segments. TEMP is used for temporary sort-related segments, and RBS is the tablespace used for rollback segments. You only want to allow developers to create segments in the DEV tablespace. Remember that they'll need to create segments in the TEMP and RBS tablespaces for normal database activities. A CREATE USER command might look like this: CREATE USER KHEIL IDENTIFIED BY KH1507 DEFAULT TABLESPACE DEV TEMPORARY TABLESPACE TEMP QUOTA 0 ON SYSTEM QUOTA 1024M ON DEV QUOTA 0 ON PROD QUOTA 0 ON TOOLS This allows no segments to be created on SYSTEM, PROD, or TOOLS. 1GB is allowed on DEV. An unlimited amount of space would be allowed on TEMP and RBS (because they aren't elementLinks/listed with any quota

http://www.informit.com/content/0789716534/element_009.shtml (6 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8

limits).

Using the CREATE SCHEMA Command

Sometimes it's useful to group several CREATE TABLE, CREATE VIEW, or GRANT commands into a single operation. You would typically do this to make sure that either everything is created properly or nothing is created. By grouping multiple CREATE TABLE/VIEW and GRANT commands into a single operation, you ensure that even if the first command succeeded, it will be rolled back if the last command fails. CREATE SCHEMA falls short of expectations Although CREATE SCHEMA offers some nice capabilities, Oracle8 doesn't support it very well. Because you can't include a STORAGE clause, you most likely can't create optimal tables with this command. Unless you need the capabilities that it offers, you'll probably get along quite well without ever using it. Suppose that you had a script with these operations: CREATE TABLE CUSTOMER CREATE TABLE SUPPLIER CREATE VIEW GRANT If the first CREATE TABLE ran successfully but the second didn't due to disk space, your database would have a CUSTOMER table but no SUPPLIER table. When the disk shortage is corrected, the script can't be used without change because it will abort when it tries to create a CUSTOMER table again. By encapsulating all these commands in a single CREATE SCHEMA transaction, you can correct the problem and restart a script if anything fails; you can do so knowing the database was returned to its original condition. The syntax of the CREATE SCHEMA command is as follows: CREATE SCHEMA AUTHORIZATION schema [CREATE TABLE ...]| [CREATE VIEW ...]| [GRANT ...] For CREATE SCHEMA AUTHORIZATION schema, provide the name of the schema where the new segments are to be created. This must be the name of the user now logged into the database. You can't use CREATE SCHEMA with someone else's schema name (or userid), even if you're a DBA. Use any valid ANSI CREATE TABLE command within a CREATE SCHEMA command. You can repeat as many CREATE TABLE commands as you want within a single CREATE SCHEMA command. Many CREATE TABLE clauses that you use, such as STORAGE, are Oracle extensions to ANSI SQL and aren't supported in the CREATE SCHEMA command.
q

Watch out for non-ANSI syntax

CREATE VIEW Just as with the CREATE TABLE command, you may use any number of valid ANSI CREATE VIEW commands. GRANT Any valid GRANT commands can be used with the CREATE SCHEMA command.

Suppose that the SUPPLIER table wasn't created due to an error, but that the CUSTOMER table was created successfully. CREATE SCHEMA wouldn't try to create the view CUSTOMERS_AND_SUPPLIERS or grant permissions. It would drop the CUSTOMER table, thereby returning the database to its original state. This is depicted in the following command:

http://www.informit.com/content/0789716534/element_009.shtml (7 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8

Used in TGASPER schema (and, of course, run by TGASPER user) Used to create CUSTOMER and SUPPLIER tables in TGASPER schema (complies with ANSI SQL) Used to create a view based on CUSTOMER and SUPPLIER tables Applies appropriate permissions on CUSTOMERS and SUPPLIERS tables

Modifying User Accounts

Like most everything else about your database, user accounts and security properties will inevitably need to be changed. Fortunately, Oracle8 provides the capability to change almost every aspect of a user account after it's created.

Using the ALTER USER Command

Oracle8 provides the ALTER USER command to change attributes associated with a user account. You can issue the ALTER USER command with as many clauses as necessary. You need to specify only the attributes you want to change; all other attributes will remain unchanged. The format of the ALTER USER command is as follows: ALTER USER userid [IDENTIFIED BY password | EXTERNALLY] [DEFAULT TABLESPACE tablespace] [TEMPORARY TABLESPACE tablespace] [QUOTA value | UNLIMITED] ON tablespace] ... [PROFILE profile] [PASSWORD EXPIRE] [ACCOUNT <LOCK | UNLOCK>] [DEFAULT ROLE <ROLE [, ROLE]... | ALL [EXCEPT ROLE [, ROLE]...] | NONE >]
http://www.informit.com/content/0789716534/element_009.shtml (8 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8
q

ALTER USER userid This will be the userid of the user you want to change. You can't rename a user after he or she is created. BY password | EXTERNALLY You can change the password of a user who isn't using external authentication. This password should begin with an alphabetical character (a-z) and consist of alphabetical or numeric characters. Be aware that this password will be visible onscreen. You should avoid changing passwords at a workstation where others could easily see the password(s) you're entering. Theoretically, you could change a user's authentication from internal (entering a userid/password) to external, or vice versa. However, because the userid will typically be different for users being authenticated internally and externally, this is rarely a practical option. You'll usually need to drop the user and recreate his or her account with the proper authentication.

Changing user authentication

q q

DEFAULT TABLESPACE tablespace You can change the default tablespace for a user at any time. Be aware that this won't move any segments already created. This affects only segments that will be created in the future if the user doesn't explicitly identify a tablespace. TEMPORARY TABLESPACE tablespace The temporary tablespace can be changed for users at any time. QUOTA value | UNLIMITED ON tablespace User quota values can be changed at any time. Just as in the CREATE USER command, you can repeat the QUOTA clause as many times as necessary. You need to provide a QUOTA clause only for the tablespaces that you want to change the user's quota on. PROFILE profile You can change a user's security profile at any time. PASSWORD EXPIRE A user's password can be expired at any time. This should always be done when a user's password is being changed by a DBA, thereby forcing the user to re-enter a password that only he or she knows. ACCOUNT LOCK | UNLOCK Accounts can be LOCKed or UNLOCKed at any time. Many sites require that when accounts are created, they must be LOCKed until the user is actually ready to use them. Use the ALTER USER command with ACCOUNT UNLOCK to unlock the user's account and allow him or her access to the database. DEFAULT ROLE You can use the DEFAULT ROLE clause to make one or more roles enabled by default on a user's account. All roles granted may be enabled by default with the ALL keyword and, optionally, exclude one or more roles from being enabled by default by explicitly elementLinks/listing them. By using the NONE keyword with the DEFAULT ROLE clause, you can specify that no roles are enabled by default. The DEFAULT ROLE clause can't be used to make (as the default) a role that a user hasn't been specifically granted with the relevant GRANT command.

DEFAULT ROLE's limitation

Example of Changing a User's Password Using ALTER USER requires that you first log into SQL*Plus as a DBA user (such as SYSTEM). The following example shows a user's password being changed:

Indicates that BJONES's user account is being changed Sets BJONES's password to BJ4236 Forces user to change password the next time she logs in (using new password) Changing User's Default Tablespace and Quota Example

http://www.informit.com/content/0789716534/element_009.shtml (9 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8

The following example shows a user's default tablespace being changed and her tablespace quotas changed accordingly:

Changes NSMITH's user account Changes user's default table-space to USERS3 Denies user any more space on tablespace USERS2 (user's old default table-space) Allows user 500MB of space on new default tablespace, USERS3 1

Understanding the Results of Changing Quota

Suppose that for the last year, JPOOLE has had a quota limit of 200MB on the USERS tablespace. His segments residing in the USERS tablespace now use around 180MB. If you were to change his quota to 100MB, there would be absolutely no effect on his existing segments-he would continue to use 180MB. He couldn't, however, use any more space. The effect of changing quota Although Oracle8 allows you to change a user's quota on any table-space at any time, doing so won't have any effect on space that has already been allocated by a user's segment(s). The only way to change the amount of space already allocated by a user in a tablespace is to copy his or her segment(s) and drop the originals from the data-base. When a segment needs to allocate an extent (initially or after it exhausts free space in existing extents), Oracle8 checks the respective user's quota against the amount of space he or she is using in the tablespace containing the segment. If allocating another extent for the user won't cause him or her to exceed the quota, Oracle8 will allocate the extent and continue. If not, the transaction will fail and the user will be notified that Oracle8 can't allocate another extent. This is the only time quotas are checked and Oracle8 takes any action based on quotas. In short, changing a quota affects only future space requests and has no effect on the space a user has already allocated to his segments.

Using the DROP USER Command

During the life of a database you'll inevitably need to remove users from the database. In Oracle8, this is accomplished with the DROP USER command. Considerations before dropping users Unfortunately, removing a user from Oracle8 has more implications than removing a user from most operating systems; a segment can't exist without an existing owner. If you must keep the segments in the doomed user's schema, you'll have to keep the schema on the data-base anyway, or move them to another user's schema. If you choose the former, lock the account with ALTER USER to prevent anyone from logging in and posing a security risk. If you want to keep a copy of the segments, you need to use Oracle8's export/import facility. The format of the DROP USER command is as follows: DROP USER userid [CASCADE] q DROP USER userid Supply the userid of the user to be dropped from the database.

http://www.informit.com/content/0789716534/element_009.shtml (10 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Creating and Managing User Accounts From: Using Oracle8
q

CASCADE If you want to remove all the segments owned by the user you're dropping, use the CASCADE option. This tells Oracle8 to first drop all the user's segments before actually dropping the user. Be careful! There's no rolling back this command.

Example of Dropping a User Account The following example shows a user being dropped, along with his objects from the database.

Drops user SCOTT from database Tells Oracle to drop any tables that SCOTT may have in his schema before actually dropping him from database If you have any doubt whatsoever about needing segments contained in a schema you're about to drop, consider exporting them to a file or tape before dropping the user. That way you can always reload the data if you need it, but your system won't be bogged down with dead wood. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_009.shtml (11 of 11) [26.05.2000 16:48:19]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

You are here : Home : Using Oracle8

Controlling User Access with Privileges

Exact Phrase All Words Search Tips

< Back

Contents

Next >

Save to MyInformIT

From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q

Levels of Access System Privileges

r r r

Granting System Privileges Revoking System Privileges Managing Privileges with ADMIN OPTION Granting Object Privileges Revoking Object Privileges Managing Privileges with GRANT OPTION Creating Roles Granting Privileges to Roles Granting Roles to Users and to Other Roles Setting Default Roles Enabling and Disabling Roles Revoking Role Privileges Dropping Roles

Object Privileges
r r r

Using Roles to Simplify Privilege Management

r r r r r r r

q q q q q

Guard your database with effective privilege management Distinguish between system and object privileges Assign privileges to different categories of users Build roles to simplify privilege management Track privilege assignments with your data dictionary

Levels of Access
Oracle protects its database access with passwords that users must know or that the system must validate. Before they can make a connection, users are still prevented from doing any work unless they have been granted the requisite privileges. Avoid using the default users Try not to connect to the database with SYS or SYSTEM any more than necessary. This way you can avoid making mistakes that can have extensive repercussions. Oracle depends on these users having the predefined characteristics with which they were created when your database was built. Instead, you should create a user to act as the primary DBA for your database, give all system privileges to that user, and allow that user to pass those privileges on to others.

http://www.informit.com/content/0789716534/element_010.shtml (1 of 19) [26.05.2000 16:48:43]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

There are generally two categories of database users: those who need to build and manage database structures and objects (such as files, tablespaces, and tables), and those who need to work with applications that use existing objects. The former users need a variety of system privileges, whereas the latter need mainly object privileges. As you might guess, system privileges allow you to create, alter, and drop various structures, whereas object privileges allow you to execute commands on specific objects. Roles allow you to group any number of system and object privileges into a named collection. When you've done this, you can administer roles in much the same way you can manage individual privileges. Typically, you use roles when you have categories of users who need the same sets of privileges to do their work. Don't take the easy way! It's not good practice to give every single privilege to anyone who needs to perform some database administrative work, even if this does appear to simplify your work. As the word "privileges" implies, you should treat them as benefits or favors, to be given only to those who have a demonstrated need and appropriate sense of responsibility. You should also ensure that users to whom you assign system privileges know how to perform the tasks that the privileges allow.

System Privileges
When you create a database, you also automatically create two users, SYS and SYSTEM. Both users are equipped with every system privilege and are allowed to add those privileges to any new users you create. The syntax for managing system privileges is relatively simple. What's much more difficult to deal with, however, is the large number of system privileges that you can assign and understanding just what each one allows (see Table 10.1). You have to decide exactly what subset of these privileges is needed by different users to complete their assigned tasks. When deciding what privileges to assign to a user, you also need to determine whether that user should be allowed to grant some or all of those privileges to other users, too. Obviously, if you're going to allow them to do this, you must be assured that they can make the same type of decision that you make when giving out the privileges in the first place-that they understand the needs and abilities of the other users to whom they'll be assigning the privileges.

Granting System Privileges

You use the GRANT command to assign system privileges to users. Each GRANT statement can assign one or more system privileges and assign the privilege(s) to one or more users. Furthermore, the command can include an option to allow recipients to grant the named privilege(s) to other users. The GRANT command can assign the privileges to the "special" user, PUBLIC, as well as to individual users. By so doing, you allow every valid user in the database to take advantage of this privilege. Realize that after you assign privileges to PUBLIC, you can't drop the privilege from individual users while retaining it for all others. Therefore, you can't use PUBLIC grants as a way to simplify the task of assigning a privilege to every user except one or two. For security reasons, system privileges don't allow users to access the data dictionary. Hence, users with ANY privileges (such as UPDATE ANY TABLE, SELECT ANY TABLE, or CREATE ANY INDEX) can't access dictionary tables and views that have not been granted to PUBLIC. The following syntax isn't complete because it doesn't include the options for managing assignments associated with roles. (These options are covered later in the section "Granting Roles to Users and to Other Roles.") This abbreviated syntax is complete for assigning individual system privileges to one or more users:

Required keywords A valid system privilege name Valid username or userid; must be included if PUBLIC isn't

http://www.informit.com/content/0789716534/element_010.shtml (2 of 19) [26.05.2000 16:48:43]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

Designates all database users; can be used alone or with a list of one or more named users (required if no username is included) Grants the right for recipient( s) to assign the named privilege(s) to, and remove it from, other users Table 10.1 lists the system privilege names that you can name in the GRANT command, as well as briefly describes what the privilege allows. Table 10.1 System privileges Privilege Name: ALTER ANY CLUSTER ALTER ANY INDEX ALTER ANY PROCEDURE ALTER ANY ROLE ALTER ANY SEQUENCE ALTER ANY SNAPSHOT ALTER ANY TABLE ALTER ANY TRIGGER ALTER ANY TYPE ALTER DATABASE ALTER PROFILE ALTER RESOURCE COST Operations Allowed to Grantee: Alter a cluster in any schema Alter an index in any schema Alter a stored procedure, function, or package in any schema Alter any role in the database Alter a sequence in any schema Alter a snapshot in any schema Alter a table in any schema Enable, disable, or compile any database trigger in any schema Alter a type in any schema Issue ALTER DATABASE commands Alter profiles

Set costs for session resources ALTER ROLLBACK SEGMENT Alter rollback segments ALTER SESSION Issue ALTER SESSION commands ALTER SYSTEM ALTER TABLESPACE ALTER USER ANALYZE ANY AUDIT ANY AUDIT SYSTEM BACKUP ANY TABLE BECOME USER COMMENT ANY TABLE CREATE ANY CLUSTER CREATE ANY DIRECTORY Privilege Name: CREATE CREATE CREATE CREATE CREATE CREATE CREATE ANY ANY ANY ANY ANY ANY ANY INDEX LIBRARY PROCEDURE SNAPSHOT SYNONYM TABLE TRIGGER Issue ALTER SYSTEM commands Alter tablespaces Issue ALTER USER commands for any user Analyze a table, cluster, or index in any schema Audit an object in any schema by using AUDIT (schema objects) commands Issue AUDIT (SQL statements) commands Export objects incrementally from the schema of other users Become another user (required to perform a full database import) Comment on a table, view, or column in any schema Create a cluster in any schema Create a directory object for BFILEs in any schema Operations Allowed to Grantee: Create an index in any schema on any table in any schema Create external procedure/function libraries in any schema Create stored procedures, functions, and packages in any schema Create a snapshot in any schema Create a private synonym in any schema Create a table in any schema Create a database trigger in any schema associated with a table in any schema Create types and type bodies in any schema (valid only with the Object option installed) Create a view in any schema Create a cluster in own schema Create a private database link in own schema Create external procedure/function libraries in own schema

CREATE ANY TYPE CREATE CREATE CREATE CREATE ANY VIEW CLUSTER DATABASE LINK ANY LIBRARY

http://www.informit.com/content/0789716534/element_010.shtml (3 of 19) [26.05.2000 16:48:43]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

Create stored procedures, functions, and packages in own schema Create a profile CREATE PROFILE CREATE PUBLIC DATABASE Create a public database link LINK CREATE PUBLIC SYNONYM Create a public synonym Create a role CREATE ROLE CREATE PROCEDURE CREATE ROLLBACK SEGMENT CREATE SEQUENCE CREATE SESSION CREATE CREATE CREATE CREATE CREATE CREATE SNAPSHOT SYNONYM TABLE TABLESPACE TRIGGER TYPE Create a rollback segment Create a sequence in own schema Connect to the database Create a snapshot in own schema Create a synonym in own schema Create a table in own schema Create a tablespace Create a database trigger in own schema Create types and type bodies in own schema (valid only with the Object option installed) Create a view in own schema Delete rows from tables or views in any schema or truncate tables in any schema Drop clusters from any schema Drop directory database objects Drop indexes from any schema Drop external procedure/function libraries from any schema Drop stored procedures, functions, or packages in any schema Drop roles Drop sequences from any schema Drop snapshots from any schema Drop private synonyms from any schema Drop tables from any schema Drop database triggers from any schema Drop object types and object type bodies from any schema (valid only with the Object option installed) Drop views from any schema Drop external procedure/function libraries Drop profiles Drop public database links Drop public synonyms Drop rollback segments Drop tablespaces Drop users Execute procedures or functions (standalone or packaged) or reference public package variables in any schema Use and reference object types, and invoke methods of any type in any schema (valid only with the Object option installed); can grant this privilege only to named users or to PUBLIC, not to a role Force the COMMIT or the rollback of any in-doubt distributed transaction in the local database or induce the failure of a distributed transaction

CREATE VIEW DELETE ANY TABLE DROP DROP DROP DROP ANY ANY ANY ANY CLUSTER DIRECTORY INDEX LIBRARY

DROP ANY PROCEDURE DROP ANY ROLE DROP DROP DROP DROP ANY ANY ANY ANY SEQUENCE SNAPSHOT SYNONYM TABLE

DROP ANY TRIGGER DROP ANY TYPE DROP ANY VIEW DROP LIBRARY DROP PROFILE DROP PUBLIC DATABASE LINK DROP PUBLIC SYNONYM DROP ROLLBACK SEGMENT DROP TABLESPACE DROP USER EXECUTE ANY PROCEDURE EXECUTE ANY TYPE

FORCE ANY TRANSACTION

http://www.informit.com/content/0789716534/element_010.shtml (4 of 19) [26.05.2000 16:48:43]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

FORCE TRANSACTION GRANT ANY ROLE INSERT ANY TABLE LOCK ANY TABLE MANAGE TABLESPACE RESTRICTED SESSION SELECT ANY SEQUENCE SELECT ANY TABLE SYSDBA

Force the COMMIT or the rollback of own in-doubt distributed transactions in the local database Grant any role in the database Insert rows into tables and views in any schema Lock tables and views in any schema Take tablespaces offline and online, and begin and end tablespace backups Connect to the database when it's running in restricted mode Retrieve numbers from sequence generators in any schema Query tables, views, or snapshots in any schema Perform Server Manager STARTUP, SHUTDOWN, and RECOVER commands, perform Server Manager ALTER DATABASE ... OPEN|MOUNT|BACKUP| [NO]ARCHIVELOG command options, and perform the CREATE DATABASE command; also includes the RESTRICTED SESSION privilege Perform Server Manager STARTUP, SHUTDOWN, and RECOVER commands, perform Server Manager ALTER DATABASE ... OPEN|MOUNT|BACKUP| [NO]ARCHIVELOG command options; also includes the RESTRICTED SESSION privilege Use an unlimited amount of any tablespace regardless of any specific quotas assigned (can grant this privilege only to named users or to PUBLIC, not to a role) Update rows in tables and views in any schema

SYSOPER

UNLIMITED TABLESPACE

UPDATE ANY TABLE

Beware of granting too much to PUBLIC System privileges should rarely, if ever, be granted to PUBLIC. Usually, at least one user shouldn't be allowed to perform the work associated with any privilege. You can't prevent such a user from exercising the privilege, however, if it's granted to PUBLIC. Privilege and usernames can't be repeated in a single GRANT command Although you can include multiple system privileges in the GRANT command, you can't name the same privilege twice. Similarly, you can't repeat a username in the list of users in the GRANT command. To grant a system privilege, you must have been granted that privilege with the ADMIN OPTION yourself, or you must have the GRANT ANY PRIVILEGE system privilege as part of your privilege domain. I recommend granting all system privileges to a DBA user with ADMIN OPTION as soon as you've built your database. You can then connect as this DBA user to perform all further system privilege administration duties unless-and until-you create other users and grant them the required privileges. For a small database, you'll probably retain these privileges just for yourself. In a large database environment, however, you may have assistant DBAs who are responsible for subsets of database administration and management. Uses for assistant DBAs You can create one or more DBA users who are solely responsible for managing the remainder of your user community, such as adding and dropping users, monitoring and assigning space, and maintaining password integrity. Similarly, you can have an assistant DBA who's in charge of space management, responsible for monitoring the availability of free space, and assigning new files or tablespaces as needed. Grant these DBA users only the system privileges necessary to perform their specific functions-the space-management DBA doesn't need user-related privileges, for example. When you grant a user a privilege, it becomes available immediately. Any user granted the privilege can begin to take advantage of its capabilities. A privilege granted with ADMIN OPTION means that these users can also assign it to or remove it from other users right away. (This includes the ability to revoke the privilege from each other and even from you or the person who granted it to them.) As said earlier, be careful to whom you give system privileges, particularly when you give them with ADMIN OPTION. To restore a system privilege that has been taken from you and to remove it from the destructive user, you need to connect as SYS and re-grant the privilege to yourself.

http://www.informit.com/content/0789716534/element_010.shtml (5 of 19) [26.05.2000 16:48:43]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

Revoking System Privileges

Use the REVOKE command to remove a system privilege from a user. If you found the syntax for the GRANT command refreshingly simple, the syntax for the REVOKE command to cancel a user's privilege(s) should be even more delightful. The syntax-again abbreviated to exclude role-related options-is as follows:

Required keywords Valid system privilege name Valid username or userid (must be included if PUBLIC isn't) Designates all database users; can be used alone or with one or more named users (required if no user-name is included) As with the GRANT command, you can't repeat a privilege name or a username in a statement. You can list as many different names as you need in each command, however, via a comma separator. In addition, as with the GRANT command, you must have the necessary privileges to revoke the named privileges. This again means that it must have been granted to you with ADMIN OPTION or you must have the GRANT ANY PRIVILEGE option. Note that there's no equivalent REVOKE privilege to equate to GRANT ANY PRIVILEGE. Revoking a privilege from PUBLIC negates the action of the GRANT command that was used to give the named privilege(s) to PUBLIC in the first place. It doesn't affect the privileges granted to individually named users; they will continue to be able to exercise any privileges not specifically revoked from them. Don't deprive yourself (of privileges) Be careful when using the REVOKE command because you can revoke a privilege from yourself if you have the necessary privileges.

Managing Privileges with ADMIN OPTION

We looked briefly at ADMIN OPTION in the preceding two sections. Recall that this option gives the recipients of the system privilege exactly the same rights you have-they not only can perform the tasks associated with the privilege (as listed in Table 10.1), but they can also assign the privilege to other users and revoke it from any user's privilege domain. I've mentioned that one user they can revoke it from is the user who granted it to them in the first place. Another point to realize about ADMIN OPTION is that when a user with this option grants the privilege to someone else, he or she can concurrently grant ADMIN OPTION for that privilege to the new user. This new user can then revoke it from their donor or from anyone else who has the privilege. The new user, of course, also has the option of granting the privilege to anyone else with or without ADMIN OPTION. It's even possible for a user to grant a system privilege to PUBLIC with ADMIN OPTION, which would allow every user to manage the privilege. The point of this discussion isn't to recommend a strategy for making privilege management easy by granting system privileges to PUBLIC with ADMIN OPTION, but rather to caution you that this can happen. Manage privileges wisely I want to emphasize the importance of managing your system privileges and ensuring that they aren't abused. Assign privileges only to the users who need them, and grant to them only the specific privileges you know they will need. Pay special attention to privileges granted with ADMIN OPTION. It is, of course, perfectly reasonable to expect that a user with a system privilege might need to give that privilege to someone else. The DBA responsible for managing your user information, for example, may need to give the CREATE, ALTER, and DROP USER privileges to a colleague in the Human Resources Department while they're away on a vacation. If you suspect that there's some misuse of system privileges, however, or you just want to monitor them as part of your security procedures, you can use the data dictionary to help. You can determine who has each system privilege, and which of the users also has the ADMIN OPTION on any given one. DBA_SYS_PRIVS is the view you need to query. A sample query against this view is as follows:

http://www.informit.com/content/0789716534/element_010.shtml (6 of 19) [26.05.2000 16:48:43]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

Name of the userid (or role, as you see later) to which the privilege has been granted Name of the granted system privilege Shows that the privilege can't be administered by the grantee Shows that the privilege can be granted or revoked by the grantee Although you can't necessarily trace who gave whom which privilege, you should be able to identify any users who have privileges you didn't expect them to have. You can begin solving your problems by removing these privileges. You may then want to remove ADMIN OPTION from the other users to prevent the privileges from being re-granted until you can investigate the users involved directly. To remove ADMIN OPTION from a system privilege, you have to revoke the privilege entirely and then re-grant the privilege without ADMIN OPTION. There isn't a command that lets you remove the administrative capability while retaining the privilege itself.

Object Privileges
Once you have a system privilege, you can use it across the whole database to do whatever the privilege allows. On the other hand, an object privilege is granted on just one object. For example, the right to query the DOCTORS table that you've been given doesn't give you any access to the PATIENTS table. This allows the owners of database tables-and other objects such as stored procedures and packages, sequences, and snapshots-to be very selective about what other users can access them. Unlike system privileges, the original database users, SYS and SYSTEM, don't own any object privileges by default, except on objects they happen to own. In fact, the owner is the only user with object privileges when an object is created. Most database objects should be accessed only through the object privileges granted by the owner to the users who have a need for them. If you've read the preceding sections on system privileges, you probably realize by now that some power users have access to all objects without being required to have any object privileges. Such users, however, should comprise a very small segment of your database user community. They need these more powerful privileges to manage the database rather than manipulate data in the applications tables. Even the owner of the tables that contain the application-related data may need only a few, if any, of the system privileges discussed earlier. In some databases I've seen, the application objects are owned by a userid that doesn't even have the privilege to connect to the database, except when an object needs to be built or modified. Protecting application schemas To protect the objects (tables, indexes, procedures, and so on) that belong to an application, create a schema to hold them. Prevent users from accidentally, or deliberately, connecting to this schema and potentially damaging its contents by with-holding the CREATE SESSION privilege. Instead, create and manage the objects for the application schema from an account with the necessary CREATE/ALTER ANY privileges. You'll temporarily have to give the application schema the ability to connect to the data-base in order to grant access privileges on the created objects. To minimize the time the account is available for logins, prepare a script to grant all object privileges WITH GRANT OPTION to the user who created the objects. All further object privileges can then be granted by this user. The typical user should access the database to perform only specific, well-regulated activities against a set of tables. For instance, a clerk in the medical office may add new patient data, record office visits, and send out statements if payment is due. To do this, he or she may need only to insert or update records in the PATIENTS table and query the ACCOUNTS_PAYABLE table. Medical technicians, on the other hand, may need to read and update the PATIENTS and INVENTORY tables and record charges in the ACCOUNTS_PAYABLE table. Both user types probably won't issue the DML commands against these tables directly, but instead use an application program that further controls their access to the records in these tables. Roles can be used to manage object privileges

http://www.informit.com/content/0789716534/element_010.shtml (7 of 19) [26.05.2000 16:48:43]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

As you see later in the "Using Roles to Simplify Privilege Management" section, you can create roles that are granted a subset of privileges, and then grant the role to the users. This minimizes the amount of privilege management on your part. Typical users require only the subset of object privileges required to complete their work on the objects that are part of the specific application they use. Although there are only nine types of object privileges (compared to about 100 different system privileges), you'll grant many more object privileges-if yours is a typical database-than you will system privileges. Similarly, most work performed in the database will be done under the permissions obtained from object privileges rather than from system privileges. You allocate and deallocate object privileges by using the GRANT and REVOKE commands, just as for system privileges. However, the two types of privileges have different characteristics and require slightly different syntax to manage them. They're also recorded in different data dictionary tables. You can query the DBA_TAB_PRIVS view to see the privileges granted on database objects:

Privilege recipient Owner of the object to which the privilege belongs Object to which the privilege belongs User who issued the grant on the privilege Name of the object privilege YES if the grantee can grant the privilege to another user, NO if the grantee can't make such a grant TERRY can grant UPDATE on SCOTT's table because GRANTABLE is YES

Granting Object Privileges

The GRANT command for object privileges has a slightly more complex structure than its system privilege counterpart: q It must identify not just the privilege, but also the object on which the privilege is being granted. q For tables and views, some privileges can be limited to a subset of the columns. q Because the objects belong to a user other than the one issuing the GRANT command, the syntax has to allow for a schema name. You can see how these factors are incorporated into the object GRANT command by looking at its syntax: GRANT [privilege_name [(column_name[,...])][,...]] [ALL [PRIVILEGES]] ON [[schema.]object_name] [DIRECTORY directory_name] TO [user_name[,...]] [PUBLIC] [WITH GRANT OPTION] GRANT, ON, and TO are the required keywords to grant object privileges. privilege_name is an object privilege. It must be included if there's no ALL clause. column_name is a column name, valid only if the privilege is named and is one of INSERT, REFERENCES, or UPDATE. ALL is used to grant all available object privileges. ALL must be included if no privilege is named. The optional keyword PRIVILEGES is used for clarification only and is never required. schema is the name of the owner of the object on which privileges are being granted. The default is the schema of the user issuing the command. object_name is the name of the object on which privileges are being granted. It must be included if there's no DIRECTORY clause. DIRECTORY directory_name identifies a directory object for storing BFILEs. It must be included if no object is named.

q q q

http://www.informit.com/content/0789716534/element_010.shtml (8 of 19) [26.05.2000 16:48:43]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8
q q

user_name is a valid username or userid and must be included if PUBLIC isn't. PUBLIC designates all database users and can be used alone or with a list of one or more named users. It's required if no username is included. WITH GRANT OPTION grants not just the named privilege(s) but also the right for the recipient(s) to assign it to, and remove it from, other users. You should be careful when granting the INSERT privilege and restricting the columns to which the privilege allows access. Any column defined with a NOT NULL constraintdirectly or indirectly through a PRIMARY KEY constraint- must be provided a value when a new row is inserted. If the INSERT privilege doesn't include all such columns, the recipient of the privilege won't be able use it because it won't allow the inclusion of values into the NOT NULL columns. SEE ALSO Learn about NOT NULL constraints,

Avoid Catch-22 when granting INSERT on selected columns

To complete the syntax options, Table 10.2 shows the names of the available object privileges and indicates on which object types they can be granted. If you use the GRANT command's ALL (PRIVILEGES) option, you can see from this table which privileges are going to be granted, based on the object type. Table 10.2 Object privileges and related objects Object Privilege: ALTER DELETE EXECUTE INDEX INSERT READ REFERENCES SELECT UPDATE Procedures Functions Packages:

Table: X X X X X X X

View: X

Sequence: X

Snapshot:

Directory:

Library:

X X X X X X X

The syntax diagram doesn't include the necessary options to grant object options to roles covered later in this chapter. To issue any object grants successfully with this command (with or without the role options), you must be the owner of the named object or have been granted the privilege(s) you're attempting to grant with GRANT OPTION. In the latter case, you need to include the schema name to identify the object's owner. Although the names of the privileges should indicate what they allow, let's quickly review each one and identify what command or commands it allows the recipient to perform. The capabilities provided by each privilege can vary depending on the type of object to which you're granting the privilege. Table 10.3 summarizes the capabilities each object privilege provides. Table 10.3 Uses for object privileges Privilege Name: ALTER DELETE EXECUTE Operations Allowed: Issue the ALTER TABLE and ALTER SEQUENCE commands Delete rows from a table or view Execute the procedure, function, package, or external procedure, and access to any program object declared in a package specification Issue the CREATE INDEX command on the table for which the privilege is given Insert rows into a table or view Read BFILEs from the named directory Create foreign-key constraints against the table

INDEX INSERT READ REFERENCES

http://www.informit.com/content/0789716534/element_010.shtml (9 of 19) [26.05.2000 16:48:44]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

SELECT UPDATE

Select rows from a table, view, or snapshot; extract numbers from a sequence generator Update rows in a table or view

Remember the ALL shortcut If you need to assign all the privileges pertinent to a specific object to one or more users, remember that you can use the ALL option in the GRANT command to grant every privilege appropriate for the object. Table 10.2 lists the privileges associated with each type of object. SEE ALSO Read more about foreign-key constraints, To help you understand object privileges, the following examples contain typical user requirements and the commands you would use to provide them: q Issue the following statement to allow PAT to query all columns in the EMPLOYEES table owned by HR, as well as to update and build indexes on them all: GRANT select, update, index ON hr.employees TO pat; Issue the following statement to allow TERRY to execute any procedures or functions in the HR package HIRE_EMPOYEE, or to access any of its publicly defined variables (see Appendix A for details): GRANT execute ON hr.hire_employee TO terry WITH GRANT OPTION; Use the following command to allow CHRIS to add new records that contain only the employee's ID number, first and last names, and birth date to the EMPLOYEES table owned by HR: GRANT insert (id, first_name, last_name, birth_date) ON employees TO chris; Issue the following statement to allow the HR_USER userid to build foreign-key constraints on the ID or SSN columns of HR's EMPLOYEES table, and to grant this right to other users: GRANT references (id, ssn) ON hr.employees TO hr_user WITH GRANT OPTION;

Revoking Object Privileges

You use the REVOKE command to remove an object privilege, just as with a system privilege. Of course, the syntax is slightly different because you need to identify a specific object. You also might have to include options to remove privileges on which other objects depend. The following syntax, as with that for the GRANT command, doesn't include the options for handling privileges assigned to roles: REVOKE [privilege_name [(column_name[,...])][,...]] [ALL [PRIVILEGES]] ON [[schema.]object_name] [DIRECTORY directory_name] FROM [user_name[,...]] [PUBLIC] [CASCADE CONSTRAINTS] [FORCE] Revoking specific column grants The REVOKE command doesn't have an option to identify individual columns. If you need to remove a privilege granted for a specific column or change an object privilege from allowing access to the whole table to restrict it to a subset of columns, you must revoke the current privilege and grant it again with the specific column limitations. All keywords and options in the REVOKE command are identical to those in the GRANT command, except the following: q REVOKE, ON, and FROM are the required keywords to revoke object privileges. q CASCADE CONSTRAINTS indicates that any referential integrity constraints defined by using the privilege being revoked will be dropped. The command will fail if such constraints exist and you don't include this optional phrase.

http://www.informit.com/content/0789716534/element_010.shtml (10 of 19) [26.05.2000 16:48:44]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8
q

FORCE automatically revokes EXECUTE privileges on user-defined object types that have table dependencies. The command will fail if such table dependencies exist and you don't include this keyword.

Unlike system privileges, which can be granted only to a user or to PUBLIC one time, an object privilege can be granted to the same user by many different grantors. The revocation of a privilege by one grantor doesn't change the status of that privilege as granted by another user to the same grantee. To remove an object privilege from a user's privilege domain or from PUBLIC, every grantor of that privilege must revoke it. Mixed PUBLIC and individual privileges If the same privilege has been granted to PUBLIC and to individual users, revoking the privilege from PUBLIC won't affect users who have been granted the privilege directly. Similarly, revoking the privilege from such a user won't preclude the use of the privilege by that user via the PUBLIC assignment. The following examples demonstrate the REVOKE command but also include GRANT commands so that you can see the whole story: q Assume that TERRY had been granted the privilege to select all columns, to update the SALARY and the JOB_TITLE columns, and to build foreign-key constraints on the DEPT column of the EMPLOYEES table owned by HR with the following command: GRANT SELECT, UPDATE (salary, job_title), REFERENCES (dept) ON hr.employees TO terry; The following commands could be used to selectively manage these privileges. Use the following to prevent further queries: REVOKE SELECT ON hr.employees FROM terry; To prevent any further definitions of foreign-key constraints on DEPT, use the following: REVOKE REFERENCES ON hr.employees FROM terry; To allow updates to the DEPT column in addition to the SALARY and JOB_TITLE columns: GRANT UPDATE (dept) ON hr.employees TO terry; To prevent updates on the SALARY column but retain it on the JOB_TITLE and DEPT columns: REVOKE UPDATE ON hr.employees FROM terry; GRANT UPDATE (job_title,dept) ON hr.employees TO terry; To remove all privileges on the table: REVOKE ALL PRIVILEGES ON hr.employees FROM terry; Combining REVOKE and GRANT commands In some situations, you may need to issue a series of REVOKE and GRANT commands to change a privilege mix on an object for a user. This is typically true when you want to remove from a user the capability to grant the privilege but want the user to retain the right to use the privilege. Automatic Cascading Impacts of Revoked Object Privileges Dependent objects may be affected by the loss of an object privilege without any further action on the part of the grantor or the grantee. This is in addition to the cascading results of the CASCADE CONSTRAINTS and FORCE options. The situations where this can occur include the following: q If you revoke a privilege on an object from a user whose schema contains a view requiring the privilege, Oracle invalidates the view. q If you revoke a privilege that's exercised in a stored procedure, the stored procedure is marked invalid. When this is done, the procedure can't be re-executed or its public variables referenced unless the privilege is regranted. (See Appendix A for details of stored procedures and public variables.) A procedure exercises an object privilege if the object is referenced by one or more SQL statements and the procedure owner isn't the object owner. q If you revoke a privilege from a user who has granted that privilege to other users, Oracle revokes the

http://www.informit.com/content/0789716534/element_010.shtml (11 of 19) [26.05.2000 16:48:44]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

privilege from grantees as well as the grantor.

Managing Privileges with GRANT OPTION

GRANT OPTION is somewhat like the ADMIN OPTION associated with system privileges. Like that option, it allows the grantee to grant the related privilege to other users. It also allows this grant to be granted with its own GRANT OPTION, thus allowing the new grantee to further grant the privilege, with or without the grant option. However, the two options behave very differently when it comes to the revocation of the privileges. Grantor object privileges can't be revoked by grantee A user who received an object privilege from a grantor who was assigned the privilege with GRANT OPTION can't revoke that privilege from the grantor, even if that user also received GRANT OPTION. (This is different from the behavior of system privileges granted with ADMIN OPTION.) To execute this REVOKE command, the grantee would require, paradoxically, the use of the privilege that the command would have to remove, due to the cascading action. And, were it allowed, the end result of this action would be that that neither the grantor nor the grantee would have the privilege. All object privileges have an ultimate owner-the owner of the object on which the grants have been made. Nobody can remove the object privileges from the object's owner, regardless of what other privileges they may possess. The object owner also has the right to determine who has any privilege on the object, regardless of how that privilege was obtained. For this reason, Oracle tracks the userids of each grantor and each grantee of every object privilege in the data dictionary. This allows the REVOKE command to cascade the removal of the privilege down to everyone who received the privilege from the user named in the REVOKE command. This can include a number of user levels, as shown in Figure 10.1. Figure 10.1 : Object privileges granted to users with the GRANT OPTION are removed with an automatic cascading REVOKE command. A user with the GRANT OPTION on an object privilege can revoke it from a user who was granted the privilege through their auspices. For example, the user HR_USER in Figure 10.1 could revoke the SELECT privilege on the EMPLOYEES table from BUDDY1 just as well as user HRUSER2. However, HRUSR1 couldn't successfully revoke the privilege from BUDDY1, even with the GRANT OPTION, because HRUSR1 didn't grant it to that user. SEE ALSO More information on user-defined objects and table dependencies,

Using Roles to Simplify Privilege Management

In a large database with many hundreds, if not thousands, of tables, stored procedures, and related objects, and with many hundreds of users, you can imagine how daunting the task of assigning privileges could be. The task is even more complicated when the database supports more than one application or many different functions within the same application, each requiring different levels of access to various objects. Within a single application or function, there can also be categories of users; some users are only allowed to browse for certain data, while others may have update rights, and yet others can have full DML capability on different subsets of tables. The arithmetic of roles and privileges Assume that you have some number of privileges (let's say m) and some number of users (let's say n) to whom privileges need to be granted. By granting each user each privilege, you have m n relationships between the privileges and the users to manage. If you use a single role, grant the privileges to the role, and then grant the role to the users, you have a total of m + n relationships to manage. The product of m and n will always be larger than their sum, with the difference getting increasingly pronounced as the magnitude increases. Consider just 100 privileges and 50 users. There would be 5,000 individual relationships if you didn't use roles-just 150 with a single role. To help you administer a database with so many different requirements for privileges, Oracle provides a special database object known as a role. Most simply, a role is a named collection of privileges. The most important characteristic of a role isn't readily apparent from this definition, however. A role can be granted just like a privilege, and the result of such a grant is that all the privileges assigned to the role will be simultaneously granted to the recipient. If you look at Figure 10.2, you can see how this reduces the number of privilege grants that you need to manage in the very simple case of four privileges being granted to three different users.

http://www.informit.com/content/0789716534/element_010.shtml (12 of 19) [26.05.2000 16:48:44]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

Figure 10.2 : With roles, you reduce the number of individual GRANT commands you issue. Roles have further powerful features: q A role can contain system and object privileges. q A role can be granted to another role, allowing you to build meta-roles. q A role can contain a mix of individual privileges plus roles, or just privileges or just roles. q A privilege granted to a role becomes immediately usable by any users assigned to the role. q Revoking a privilege from a role immediately prevents any further use of the privilege by all users associated with the role. q A role can be in an activated or a deactivated state for a user. q A role can be password-protected to prevent users from activating its privileges for themselves. Activated roles By allowing roles to be active or inactive, Oracle provides an extra level of security for the privileges managed by the role. Until the role is active for a user, that user can't take advantage of the privileges associated with the role. If the role is password protected, the user can't activate it without the password. In such cases, the role can be activated by an application program that can supply the password without the user having knowledge of it. This prevents accidental or deliberate use of the privileges while a user is connected to the database through some means other than the application.

Creating Roles
Anyone with the requisite CREATE ROLE privilege can create a new role. As soon as it's created, the role doesn't belong to that user's schema, nor to any other schema. To do anything with a role, however, you must have privileges on it, and only the original creator automatically receives the equivalent of a system privilege's ADMIN OPTION on a role. If the original creator wants to allow other users to administer the role, he or she can grant it to those users with ADMIN OPTION just as though it were a system privilege. Just as with system privileges, when users have ADMIN OPTION on a role, they can grant the role to others as well as revoke it from other users-including the original creator. Maintaining the role administration function At least one user must maintain the ADMIN OPTION on a role at all times. As with system privileges, you can't revoke just the ADMIN OPTION-you must revoke all access to the role. Unlike system privileges, you can't revoke the role from yourself; however, you can remove the role from the data-base if you have no further use for it, which, of course, eliminates the need for it be administered any further. The command for creating a role has the following syntax: CREATE ROLE role_name [NOT IDENTIFIED] [IDENTIFIED [BY password] [EXTERNALLY] [GLOBALLY]] CREATE ROLE is the command name. NOT IDENTIFIED indicates that the role can be enabled without a password. This is the default if the IDENTIFIED clause isn't included, so this clause can be omitted. IDENTIFIED indicates that further authorization is required to enable the role. No such authorization will be required if this option is omitted. BY password names the password that the user must supply to enable the role. It must be included in the IDENTIFIED clause if the EXTERNALLY and GLOBALLY options aren't. EXTERNALLY indicates that the operating system, or a security service running under it, will provide authentication to enable the role. External authentication may or may not require a password. This keyword must be included in the IDENTIFIED clause if the BY and GLOBALLY options aren't. GLOBALLY indicates that the user must be authorized by the Oracle Security Service before the role can be enabled. (See The Oracle Security Server Guide and the Oracle8 Server Distributed Systems manual for further details.) This keyword must be included in the IDENTIFIED clause if the BY and EXTERNALLY options aren't.

q q

External authentication of roles

http://www.informit.com/content/0789716534/element_010.shtml (13 of 19) [26.05.2000 16:48:44]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

Only certain operating systems provide facilities for external authentication of roles. You must check your operating system-specific documentation for information on this topic, including the steps needed to set up specific authorizations. You can also use third-party products for external authentication of roles. Consult the vendor's documentation and The Oracle Security Server Guide for more details on such products. You can change the authorization required for a role after it's created by issuing an ALTER ROLE command. This command accepts the NOT IDENTIFIED or IDENTIFIED clause, the latter requiring one of three options listed in the CREATE ROLE syntax. However, before you can alter a role to global authorization, you must revoke the following: Privileges for altering a role To alter a role, you must have ADMIN OPTION on the role or the ALTER ANY ROLE system privilege. If you're altering the authentication from global to unrestricted (NOT IDENTIFIED) or to alternate authentication (BY password or EXTERNALLY), you'll be automatically granted ADMIN OPTION on the role if you don't already have it.
q q q q

All other roles granted to the role and identified externally The role from all users, except the one altering the role The role from all other roles The role from PUBLIC

Oracle provides this set of predefined roles with included privileges when the database is created: q CONNECT provides basic privileges for a typical application user. q RESOURCE provides basic privileges for an application developer. q DBA provides a full set of system privileges with ADMIN OPTION. q EXP_FULL_DATABASE and IMP_FULL_DATABASE are for users of the Export and Import utilities. Ghost roles A number of roles are added to the database when it's created that don't have any associated privileges until work is done that requires them. These roles are undocumented and are included here for completeness: EXECUTE_CATALOG_ ROLE, SELECT_CATA-LOG_ ROLE, SNMPAGENT, DELETE_CATALOG_ROLE, AQ_USER_ROLE, RECOV-ERY_ CATALOG_OWNER, and AQ_ADMINISTRA-TOR_ ROLE. Oracle doesn't recommend relying on the CONNECT, RESOURCE, or DBA roles, which are included for compatibility with earlier versions of the database and may not be included in later releases.

Granting Privileges to Roles

You grant system and object privileges to roles by using the GRANT commands discussed earlier. The syntax is the same as discussed under each command format, except that you use a role name rather than a username. There's also one restriction on object privileges granted to roles-you can't use WITH GRANT OPTION. To grant privileges to a role, you must either have the GRANT ANY ROLE system privilege or have ADMIN OPTION on the role, either by being its creator or by being granted the role with this privilege. Grant assignee options In a single GRANT command, you can grant privileges to a user, a list of users, a role, a list of roles, or a list composed of users and roles. The list can even include users who are already assigned to one or more of the roles in the list. PUBLIC is also a valid user in this context. As soon as a privilege is granted to a role, it's usable to anyone with the role enabled, and will also be available to anyone enabling the role in the future. There are certain restrictions on the privileges associated with roles. Objects that rely on other object's privileges can't be created when those privileges are available only through a role rather than be directly granted. For example, a view can't be created against a table if the SELECT privilege on that table is available only through a role.

Granting Roles to Users and to Other Roles

A role is granted to a user or to another role by using the same syntax as the system privilege GRANT command. This syntax, explained in detail earlier in the section "Granting System Privileges," can be followed exactly except that the name of the role is substituted for the name of a system privilege. A single GRANT command can
http://www.informit.com/content/0789716534/element_010.shtml (14 of 19) [26.05.2000 16:48:44]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

include multiple role names as well as one or more system privilege names. As mentioned earlier, you can't repeat a name in this list. When a role is granted to another role, the privileges from both roles are combined. The granted role will pass on this superset of privileges to anyone who is granted the role. If this role is itself granted to another role, the privileges from the two original roles will be assigned to the newly granted role; see Figure 10.3 for an illustration of this. Figure 10.3 : You can build complex roles from more simple roles. Most application users will have a fixed set of tasks that they need to complete on an ongoing basis. Many of these users will, in fact, perform identical tasks to each other. A single role can be used to provide these users with all the necessary system and object privileges to perform their work. Occasionally, these users may need to perform special projects. A different role can be created to handle the associated privileges and granted to the base role for the duration of the project. One or two users may need both the base role and the special role on a regular basis, in which case you can grant both roles to these users. Other users may perform multiple tasks in a given application. These users can also be assigned their privileges through a role, but in this case the role will contain the base roles from each independent task. These users' supervisors may have yet another role that contains extra privileges for their supervisory functions, plus the complex role already created for their staff. If you have a situation similar to this, you should plan your roles carefully so that you're allowed the most flexibility in assigning them to the different worker categories. By this time, you should be able to follow the full syntax of the GRANT command for system privileges and roles. It's shown here without further comment. If you aren't clear on any of the options, refer to the explanation earlier in this chapter. GRANT [privilege_name][,...] [role_name][,...] [,...] TO [user_name[,...]] [role_name][,...] [,...] [PUBLIC] [WITH ADMIN OPTION] If the GRANT command includes the WITH ADMIN OPTION clause, any role being granted by the command will be available to the grantees just as if the role had been granted to them directly. In other words, the grantees have the right to grant and revoke the role to and from other users as well as the right to grant and revoke other roles and privileges to and from the role. There are some limitations on grants associated with roles. In particular, you can't grant the following: q A global role to a user or to any other role q A role IDENTIFIED EXTERNALLY to a global user or to a global role q A global role to PUBLIC q A role to itself, directly or through a circular set of grants Avoid granting roles circularly A circular grant would occur if Role A were granted to Role B, Role B were granted to Role C, and Role C were granted to Role A. This would result in a grant of Role A to itself. Oracle doesn't support granting a role to itself, however, and will prevent you from doing this accidentally by issuing an error message if you try.

Setting Default Roles

A role is automatically active for a user to whom it's granted, even if it's a passworded role. To restrict access to the privileges awarded through a role, you have to change the privilege domain of each granted user for whom you don't want the role active by default. You have to use the ALTER USER command's DEFAULT ROLE option to do this. In a single statement, you can identify exactly which roles granted to the user you want active or inactive whenever the user first connects to the database. Why identify default roles? As noted, all granted roles are active by default whenever a user logs into the database. You don't want these roles enabled all the time if you've created roles with password protection to prevent users from using the related privileges unless under the control of an application. The only way to prevent them from being active is to identify all the roles that should be active in the ALTER USER command's DEFAULT ROLE clause. Any role not named, or any role named in the ALL option's EXCEPT clause, will remain disabled.

http://www.informit.com/content/0789716534/element_010.shtml (15 of 19) [26.05.2000 16:48:44]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

The syntax for only the DEFAULT ROLE option of the ALTER USER command is as follows: ALTER USER user_name DEFAULT ROLE [role_names] [ALL [EXCEPT role_names]] [NONE] ALTER USER user_name identifies the user's name. DEFAULT ROLE is the required option. role_names is the name of a single role, or a comma-separated list of role names, which should be active at connect time. ALL indicates that all granted roles are to be active at connect time. EXCEPT role_names identifies a single role, or a comma-separated list of role names, that shouldn't be active at connect time. This phrase can be used only with the ALL option. NONE causes no roles to be active at connect time.

q q q

q q

Password versus external or global authentication of roles You should use password protection for roles that assign privileges you never want the user to be able to take advantage of directly. The privileges granted by the role are used only inside programs that use SQL to activate the role when needed- supplying the password, of courseand deactivate it before exiting. External and global authentication offer a layer of security for roles as long as the operating system account isn't compromised. Adding passwords to the roles gives an added layer of security, but users must know the passwords to activate them at any time. Passworded roles using external or global authorization can't be activated from a program. You must include one-and only one-of the role_name (or list), ALL, or NONE options in a single command. You can't enable roles with the ALTER USER command if any of the following apply: q The role hasn't been granted to the user. q The role was granted to another role (unless it was also granted to the user directly). q The role is managed by an external service or by the Oracle Security Service Certification Authority. Here are some examples of the commands you would use to set up the default roles for different users: q To activate only the ACCT_PAY role for Kim: ALTER USER kim DEFAULT ROLE acct_pay; To activate all but the MONTH_END role for Pat: ALTER USER pat DEFAULT ROLE ALL EXCEPT month_end; To prevent Terry from having any active roles at connect time: ALTER USER terry DEFAULT ROLE NONE; SEE ALSO A detailed description of the ALTER USER command,

Enabling and Disabling Roles

To activate a role that has been granted to you but isn't available to you now, you have to issue the SET ROLE command and identify the role you need enabled. This will generally be done inside an application, where the role can be enabled without you having to know its password. The SQL command to enable a role is as follows: SET ROLE [role_name [IDENTIFIED BY password][,...]] [ALL [EXCEPT role_name[,...]] [NONE] Using the ALL EXCEPT clause

http://www.informit.com/content/0789716534/element_010.shtml (16 of 19) [26.05.2000 16:48:44]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

You can't use the ALL option if one or more of your roles requires a password to enable it. If even just one of them requires a password, you'll have to list each role-including the password when necessary-in the SET ROLE command. The EXCEPT clause allows you to use the ALL option by excluding the role or roles with passwords. Notice that this command is very similar to the ALTER USER command's DEFAULT ROLE clause discussed in the previous section. The only differences are the keywords SET ROLE and the optional IDENTIFIED BY password clause. The latter is required to activate a password-protected role. As with the ALTER USER command, you must include one of three optional clauses-role name or list, ALL, or NONE-in the SET ROLE command. Because the SET ROLE command works only for the user who issues it, there's no provision to include a username. The SET ROLE command disables roles SET ROLE disables all roles for the user-including those identified as default roles- except those specifically enabled by the command. Before issuing this command, you should identify any other roles you may need to remain active along with the additional ones you're trying to enable. Include all these roles in the SET ROLE command by naming them or by using the ALL option. Another similarity between the SET ROLE command and the ALTER USER...DEFAULT ROLE command is that the statement overrides the default behavior. The command activates only the named roles or all the roles not listed in the ALL option's EXCEPT clause. All other roles are disabled by default. In some cases, the application that needs to change your active roles will be executing a PL/SQL routine rather than a program that can issue the SET ROLE command. To accommodate this eventuality, Oracle provides the procedure SET_ROLE as part of the DBMS_SESSION package. The procedure accepts one string as an input parameter; this string should contain a valid SET ROLE command option, just as shown in the preceding syntax. The following listing shows an anonymous PL/SQL block that activates the ACCT_PAY and MONTH_END roles: BEGIN DBMS_SESSION.SET_ROLE('ACCT_PAY,MONTH_END'); END; The DBMS_SESSION procedure can't be used in a stored procedure or trigger, and the role(s) it activates may not be accessible until after the PL/SQL block completes successful execution.

Revoking Role Privileges

The REVOKE command is used to revoke privileges from roles, roles from roles, and roles from users or from PUBLIC. To revoke an object or a system privilege from a role, use the appropriate version of the REVOKE as discussed earlier, substituting a role name anywhere a username can be used. You shouldn't need the CASCADE CONSTRAINTS or FORCE clause for object privileges because no objects can be created by using object privileges granted via a role. If you revoke a role to which other roles have been granted, the entire set of privileges associated with every role will be revoked. Of course, if any of those roles and privileges had been granted directly to a user or a role affected by the revoke, they can still exercise the related privileges through the direct grant.

Dropping Roles
You can drop a role at any time as long as you have the role granted to you with ADMIN OPTION or have the DROP ANY ROLE system privilege. The DROP command takes effect immediately, revoking the role from all granted users and other roles and then removing it from the database. The command is as follows: DROP ROLE role_name It has no options. Any roles granted to the dropped role will remain defined but won't be available for use by anyone unless they have been granted those roles directly. Data Dictionary Tables and Privilege Tracking The data dictionary uses a number of tables to manage all the possible relationships due to the complexity that

http://www.informit.com/content/0789716534/element_010.shtml (17 of 19) [26.05.2000 16:48:44]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

can result from the granting of multiple privileges to single users and to roles, and then the granting of roles to users and to other roles. To help you and your users keep track of what has been assigned to whom and to what, and what is or isn't currently active, Oracle provides a number of views on these data dictionary tables. Table 10.4 briefly summarizes the data dictionary views related to privileges and roles. Table 10.4 Data dictionary views to monitor privileges and roles View ALL_COL_PRIVS ALL_COL_PRIVS_MADE ALL_COL_PRIVS_RECD ALL_TAB_PRIVS ALL_TAB_PRIVS_MADE ALL_TAB_PRIVS_RECD Description Shows grants on columns for which the user or PUBLIC is the grantee Shows grants on columns for which the user is the owner or the grantor Shows the grants on columns for which the user or PUBLIC is the grantee Shows grants on objects for which the user or PUBLIC is the grantee Shows grants on objects for which the user is the owner or the grantor

Shows the grants on objects for which the user or PUBLIC is the grantee Shows grants on columns for which the user is the owner, grantor, COLUMN_PRIVILEGES or grantee, or PUBLIC is the grantee Shows all grants on columns in the database DBA_COL_PRIVS DBA_PRIV_AUDIT OPTIONS Shows all system privileges being audited Shows all roles in the database DBA_ROLES DBA_ROLE_PRIVS DBA_SYS_PRIVS DBA_TAB_PRIVS HS_EXTERNAL_OBJECT PRIVILEGES HS_EXTERNAL_USER PRIVILEGES ROLE_ROLE_PRIVS ROLE_SYS_PRIVS ROLE_TAB_PRIVS SESSION_PRIVS SESSION_ROLES SYSTEM_PRIVILEGE MAP TABLE_PRIVILEGES TABLE_PRIVILEGE_MAP USER_COL_PRIVS USER_COL_PRIVS_MADE USER_COL_PRIVS_RECD USER_ROLE_PRIVS USER_SYS_PRIVS USER_TAB_PRIVS USER_TAB_PRIVS_MADE USER_TAB_PRIVS_RECD Shows the roles granted to users and to other roles Shows system privileges granted to users and to roles Shows all grants on objects in the database Shows information about privileges on non-Oracle data stores Shows information about granted privileges that aren't tied to any particular object related to non-Oracle data stores For roles to which the user has access, shows roles granted to other roles For roles to which the user has access, shows system privileges granted to roles For roles to which the user has access, shows object privileges granted to roles Shows privileges now available to the user Shows roles now available to the user Maps system privilege names to privilege codes numbers Shows grants on objects for which the user is the owner, grantor, or grantee, or PUBLIC is the grantee Maps object privilege names to privilege codes Shows grants on columns for which the user is the owner, grantor, or grantee Shows the grants on columns of objects owned by the user Shows the grants on columns for which the user is the grantee Shows the roles granted to the user Shows the system privileges granted to the user Shows the grants on objects for which the user is the owner, grantor, or grantee Shows the grants on objects for which the user is the owner Shows the grants on objects for which the user is the grantee

http://www.informit.com/content/0789716534/element_010.shtml (18 of 19) [26.05.2000 16:48:44]

informit.com -- Your Brain is Hungry. InformIT - Controlling User Access with Privileges From: Using Oracle8

Views for compatibility The COLUMN_PRIVILEGES and TABLE_PRIVILEGES views are provided only for compatibility with earlier versions. Oracle recommends avoiding the use of these views. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_010.shtml (19 of 19) [26.05.2000 16:48:45]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Auditing Database Use and Controlling Resources and Passwords

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

Why Audit?
r r r r r

Preparing the Audit Trail Maintaining the Audit Table Controlling System Auditing Controlling Object Auditing Reviewing Audit Records

q q

Profiles and System Resources Combined Resource Limits

r r r r

Creating Profiles Assigning Profiles Altering Profiles Dropping Profiles Creating Password-Management Profile Entries Checking for Password Complexity

Profiles and Password Management

r r

q q q q

Control what's audited Interpret audit results Control consumption of system resources Enforce password management

Why Audit?
For an Oracle DBA, auditing is the process of recording what's being done in the database. Audit records can tell you what system privileges are being used and how often, how many users are logging during various periods, how long the average session lasts, what commands are being used against specific tables, and many other related facts. However, you shouldn't treat auditing as an idle gatherer of data-you simply add unnecessary overhead by doing that. You should use auditing when it's the easiest, fastest, and least intrusive means of collecting information that you need to perform your job. Why can't you audit SYS?

http://www.informit.com/content/0789716534/element_011.shtml (1 of 18) [26.05.2000 16:49:10]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

Oracle won't let you audit the SYS userid to keep the database from reaching some type of resource limit that needs the SYS userid to correct. If one or more of the steps required to fix the problem is audited and the problem being addressed is interfering with the system's capability to create audit records successfully, SYS can't complete the necessary corrective steps. The types of activities you can perform as a DBA where auditing could help include the following: q Preparing database usage reports for your management (how many users connect per day/week, how many queries are issued per month, or how many employee records were added or dropped last week) q Recording failed attempts to break into the database if you suspect hackers q Identifying the busiest tables that may need additional tuning q Investigating suspicious changes to critical tables q Projecting resource consumption by an anticipated increase in user load Depending on your requirements, you should be able to target exactly what level of auditing you need to invoke. In some cases, you may need to begin by gathering general data before you can identify critical areas that need closer attention. You may suspect, for example, that the fragmentation in your index tablespace is the result of excessive index creation and deletion over the weekends. You don't know with which tables the indexes are associated because they don't stay in place long enough to find them. You can audit the general activity of index creation in order to identify the specific table or tables involved. You can then audit the indexing activity of the users who have permission to add and drop indexes on these tables. From these detailed audit records, you may be able to identify which user is causing the problem, or at least the specific times the activity is occurring. You can then run a program or log in yourself during these periods to capture data dictionary information about the indexes while they're still in place. Targeting the problem If problems can be resolved by carefully honing in on the most specific details from the more general ones, you need to audit only for the exact level of detail you need at each step. This reduces the amount of extraneous data that the audit routines will capture and that you have to filter out to get to what you really need.

Preparing the Audit Trail

When you decide you need to audit, you must configure your database to handle the task. This requires a special entry in your parameter file. The audit task can be activated-and similarly deactivated-only when the database instance is stopped and restarted. Protect your audit trail If you're auditing for any reason- particularly if you're suspicious of hackers-you should make sure that you protect your audit trail from potential tampering. If you're using an external file, make sure that it's in a protected directory. If you're using the internal AUD$ table, don't give privileges on the table to any user unnecessarily. Just to make sure that you have a record of any attempts to tamper, audit any activity, DDL or DML, against the audit trail by using the commands you learn in the upcoming sections. The parameter you need to modify is AUDIT_TRAIL. By default, it has a value of NONE, which disallows any auditing regardless of any other auditing options that may have been set. You have two additional choices of setting for AUDIT_TRAIL-DB or OS-although the values TRUE (corresponding to DB), and FALSE (corresponding to NONE), are also supported for backward compatibility. With AUDIT_TRAIL set to DB, the audit records are written to a database table named AUD$, which is owned by SYS. If set to OS, the audit records are written to an external operating system file. You can control the location of the audit file by setting an additional parameter, AUDIT_FILE_DEST, to identify the full path name of the directory where you want it to be written. By default, the AUD$ table is built in the SYSTEM tablespace. This can cause problems if you do extensive auditing because it's possible to fill up this tablespace very quickly. The number of audit records created depends on just how many different actions you audit and what level of detail you require. You should be prepared to monitor your audit table or file on regular basis, even if you collect only minimal data, because other users may turn on additional auditing (by design or by accident), which can generate far more data than you were anticipating.

http://www.informit.com/content/0789716534/element_011.shtml (2 of 18) [26.05.2000 16:49:10]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

Maintaining the Audit Table

Although you're never supposed to modify the contents or the definition of the tables in the data dictionary (meaning any objects owned by SYS in the SYSTEM tablespace), this restriction is removed from the AUD$ table. You can change its storage parameters for it to grow larger than set by default, or to make better use of the free blocks in its tablespace. You also may need to remove old or other unwanted records to reclaim space. You can even copy records you do want to analyze to some other location (including an external file) and then drop them from the AUD$ table. In fact, doing this on a regularly scheduled basis prevents the table from growing uncontrollably. Controlling the capability to perform auditing There are two levels of auditing: on statements and on objects. You can control the auditing of statements by granting only the AUDIT SYSTEM privilege to users whom you want to be able to perform this type of auditing. Object auditing is a little more difficult to control because an object's owner can request that auditing be performed against the use of that object. However, you should be able to prevent any unwanted auditing if you allow objects to be created only by a restricted set of users and protect the schemas where those objects are created by removing connection privileges to them. If the objects do need to be audited, you can grant the AUDIT ANY system privilege to the user(s) responsible for this task and remove it again when the required auditing is established. You can store records from the AUD$ table a number of ways: q Issue the SQL statement CREATE TABLE audit_day1 AS SELECT * FROM sys.aud$. This will create a table called AUDIT_DAY1 and copy all the records from AUD$ into it. q Run the Export utility as user SYS and select the table AUD$ to export. q Use the SQL*Plus SPOOL command and execute a query against the sys.aud$ table or against an audit view. (See "Reviewing Audit Views" later in this chapter.) After you save the records you need from AUD$, you can remove the records you no longer need via different commands, depending on what you're trying to accomplish. Use the following command to remove all records from the table: DELETE FROM sys.aud$; Execute the following command to remove all records and return the table to its initial extent size: TRUNCATE TABLE sys.aud$; Use the following to remove all records associated with the user Kim: DELETE FROM sys.aud$ WHERE userid = 'KIM'; You can issue the following command to remove all records except those associated with the EMPLOYEES table: DELETE FROM sys.aud$ WHERE obj$name <> 'EMPLOYEES'; SEE ALSO Discover the specifics of starting and stopping a database, An introduction to the parameter file,

Controlling System Auditing

You can audit the use of system privileges or the use of SQL statements used to either administer the database or to manage database objects with a pair of commands. You can also audit the use of DML and DCL statements on specific objects by using a slightly different version of these commands (described in the next section). In both cases, the two commands are AUDIT and NOAUDIT; the former begins an audit and the latter terminates it. Whether you're auditing at the database or object level, the AUDIT command won't generate any audit records unless the instance is running with auditing enabled, as described in the previous section. The AUDIT command, which begins auditing on system privileges or database administrative commands, has the following syntax:

http://www.informit.com/content/0789716534/element_011.shtml (3 of 18) [26.05.2000 16:49:10]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

q q

AUDIT [system_privilege] [,...] [statement_option] [,...] [shortcut_name] [,...] [BY [user_name [,...]]] [BY [SESSION] [ACCESS]] [WHENEVER [NOT] SUCCESSFUL] AUDIT is the keyword to turn on auditing. system_privilege [,...] is a system privilege or a comma-separated list of system privileges. You can audit any of the privileges listed in Table 10.1. You must include at least one system privilege if you don't list any statement options or shortcut names. Although you can't audit the user SYS yourself, Oracle does track the key work done by SYS or using SYS's privileges. This auditing is written to a special file that varies by name and location depending on your operating system. Instance start-up, instance shutdown, and connection as administrator are among the key items audited. SEE ALSO Table 10.1 lists privileges on statement_option [,...] is a statement option or a comma-separated list of statement options from Table 11.1. You must include at least one statement option if you don't list any system privileges or shortcut names. shortcut_name [,...] is a shortcut name or a comma-separated list of shortcut names from Table 11.2. You must include at least one shortcut name if you don't list any system privileges or statement options. BY user_name [,...] identifies a user, or a list of users, on whom the auditing of the chosen actions will be performed. All users will be audited if you omit this option. BY SESSION/ACCESS determines whether all use of an audited action is summarized in a single audit record per user session (BY SESSION), or if a separate audit record will be generated each time the action is performed (BY ACCESS). Audited actions will be summarized at the session level if you omit this option. WHENEVER [NOT] SUCCESSFUL determines whether only successful (SUCCESSFUL) or unsuccessful (NOT SUCCESSFUL) attempts to use the audited action are recorded. If you omit the WHENEVER option, all actions-successful and unsuccessful-are audited. Even if you audit at the session level (by default or with the BY SESSION option), you can still generate access-level audit records. This is because all Data Definition Language (DDL) statements, audited as a result of selected system privileges or statement options, will be audited by access.

Default auditing of the SYS user

Default auditing for DDL statements

Table 11.1 Statement options for the AUDIT command Statement Option: ALTER SEQUENCE/1/ ALTER TABLE/1/ CLUSTER COMMENT TABLE/1/ Audited SQL Statements and Operations: ALTER SEQUENCE ALTER TABLE CREATE CLUSTER, AUDIT CLUSTER, DROP CLUSTER, TRUNCATE CLUSTER COMMENT ON TABLE for tables, views, or snapshots COMMENT ON COLUMN for table columns, view columns, or snapshot columns CREATE DATABASE LINK, DROP DATABASE LINK DELETE FROM tables or views CREATE DIRECTORY, DROP DIRECTORY Execution of any procedure or function, or access to any variable, library, or cursor inside a package GRANT privilege ON directory REVOKE privilege ON directory

DATABASE LINK DELETE TABLE/1/ DIRECTORY EXECUTE PROCEDURE/1/

GRANT DIRECTORY/1/

http://www.informit.com/content/0789716534/element_011.shtml (4 of 18) [26.05.2000 16:49:10]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

GRANT PROCEDURE/1/ GRANT SEQUENCE/1/ GRANT TABLE/1/ GRANT TYPE/1/,/2/ INDEX INSERT TABLE/1/ LOCK TABLE/1/ NOT EXISTS PROCEDURE

GRANT privilege ON procedure, function, or package REVOKE privilege ON procedure, function, or package GRANT privilege ON sequence REVOKE privilege ON sequence GRANT privilege ON table, view, or snapshot REVOKE privilege ON table, view, or snapshot GRANT privilege ON TYPE REVOKE privilege ON TYPE CREATE INDEX, ALTER INDEX, DROP INDEX INSERT INTO table or view LOCK TABLE table or view Any SQL statement failures because referenced objects don't exist CREATE FUNCTION, CREATE LIBRARY, CREATE PACKAGE, CREATE PACKAGE BODY, CREATE PROCEDURE, DROP FUNCTION, DROP LIBRARY, DROP PACKAGE, DROP PROCEDURE CREATE PROFILE, ALTER PROFILE, DROP PROFILE CREATE PUBLIC DATABASE LINK, DROP PUBLIC DATABASE LINK CREATE PUBLIC SYNONYM, DROP PUBLIC SYNONYM CREATE ROLE, ALTER ROLE, DROP ROLE, SET ROLE CREATE ROLLBACK SEGMENT, ALTER ROLLBACK SEGMENT, DROP ROLLBACK SEGMENT Any statement containing a sequence.CURRVAL or sequence.NEXTVAL phrase, where sequence is the name of an Oracle sequence generator SELECT FROM table, view, or snapshot CREATE SEQUENCE, DROP SEQUENCE All database logins CREATE SYNONYM, DROP SYNONYM AUDIT, NOAUDIT GRANT, REVOKE CREATE TABLE, DROP TABLE, TRUNCATE TABLE CREATE TABLESPACE, ALTER TABLESPACE, DROP TABLESPACE CREATE TRIGGER ALTER TRIGGER with ENABLE and DISABLE options DROP TRIGGER ALTER TABLE with ENABLE ALL TRIGGERS and DISABLE ALL TRIGGERS clauses CREATE TYPE, CREATE TYPE BODY, ALTER TYPE, DROP TYPE, DROP TYPE BODY UPDATE table or view CREATE USER, ALTER USER, DROP USER CREATE VIEW, DROP VIEW

PROFILE PUBLIC DATABASE LINK PUBLIC SYNONYM ROLE ROLLBACK STATEMENT SELECT SEQUENCE

SELECT TABLE/1/ SEQUENCE SESSION SYNONYM SYSTEM AUDIT/3/ SYSTEM GRANT/4/ TABLE TABLESPACE TRIGGER

TYPE/2/ UPDATE TABLE/1/ USER VIEW

/1/ /2/ /3/ /4/

Not included in the ALL shortcut Available only with the Object option When used with system privileges or statement options When used with system privileges and roles

To save you from having to enter a series of related system privileges or statement options in an AUDIT statement, Oracle provides a series of shortcuts. Each of these, when referenced in an AUDIT statement, causes auditing to occur on the related items. Table 11.2 lists these shortcuts and the system privileges and statement options included when you use them in an audit command. Not all AUDIT options can be named in a single statement

http://www.informit.com/content/0789716534/element_011.shtml (5 of 18) [26.05.2000 16:49:11]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

You can mix statement options and shortcuts in the same statement when issuing the AUDIT command, but you can't include most system privileges with statement options or shortcuts. Other than this restriction, you can include as many auditing choices as you want in a single statement. You can even include a shortcut and the statement options covered by the short-cut. Similarly, you can either include as many users as you want or allow the statement to default to all users. Table 11.2 Shortcuts for system privileges and statement options Shortcut Name: CONNECT RESOURCE System Privilege (P) or Statement Option (O): Privileges and Options Included: P CREATE SESSION P P P P P P P P P P P P P P P P O O P ALTER SESSION CREATE CLUSTER CREATE DATABASE LINK CREATE PROCEDURE CREATE ROLLBACK SEGMENT CREATE SEQUENCE CREATE SYNONYM CREATE TABLE CREATE TABLESPACE CREATE VIEW AUDIT SYSTEM CREATE PUBLIC DATABASE LINK CREATE PUBLIC SYNONYM CREATE ROLE CREATE USER SYSTEM GRANT All statement options listed in Table 11.1, except those noted as not being part of the ALL shortcut All system privileges

DBA

ALL ALL PRIVILEGES

Use the NOAUDIT command to cease auditing of the actions you defined with the AUDIT command. The syntax is identical to the AUDIT command except that there's no BY SESSION or BY ACCESS option; the NOAUDIT option turns off whatever option is in effect. The NOAUDIT command's syntax is as follows: NOAUDIT [system_privilege] [,...] [statement_option] [,...] [shortcut_name] [,...] [BY [user_name [,...]]] [WHENEVER [NOT] SUCCESSFUL] The various options are described in the syntax description for the AUDIT command. You can use the NOAUDIT command to stop auditing successful or unsuccessful actions if your AUDIT command had enabled both options (the default) when you executed it. However, you can't alter the auditing behavior from successful to unsuccessful with this command; you have to disable the auditing and re-enable it with your preferred option by issuing a new AUDIT command. If, on the other hand, you were auditing only successful or only unsuccessful actions, the NOAUDIT command can turn off the auditing if you issue it with no WHENEVER option. The NOAUDIT default doesn't necessarily stop all auditing If your NOAUDIT command doesn't include a list of users, auditing enabled by an AUDIT command also issued without a user list will be terminated. However, any users being audited for the same action(s) for whom you turned on auditing (by naming them in an AUDIT command) won't be affected. Turning off auditing with the NOAUDIT command won't affect any records already created as a result of the previous AUDIT command, but will prevent any further audit records from being created as a result of the

http://www.informit.com/content/0789716534/element_011.shtml (6 of 18) [26.05.2000 16:49:11]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

audited actions. Let's end this section by looking at some examples of how the AUDIT and NOAUDIT commands are used: q The following three commands, in turn, activate the auditing of all connections to the database, the auditing of any successful attempt by Kim to use the ALTER ANY TABLE system privilege, and any attempt by other users to use this privilege: AUDIT create session; AUDIT alter any table BY kim WHENEVER SUCCESSFUL AUDIT alter any table; To audit just unsuccessful connection attempts, you could now issue this command: NOAUDIT create session WHENEVER SUCCESSFUL; To audit just successful attempts by all other users (as well as Kim) to use the ALTER ANY TABLE privilege, you could issue this command: NOAUDIT alter any table WHENEVER NOT SUCCESSFUL; To stop the auditing of successful uses of the ALTER ANY TABLE privilege by everyone, you would enter either one of these two statements: NOAUDIT alter any table; NOAUDIT alter any table WHENEVER SUCCESSFUL; However, Kim's successful use of this privilege would still be audited due to the separate AUDIT command issued earlier. To discontinue this auditing, you also need these two commands: NOAUDIT alter any table BY kim; NOAUDIT alter any table BY kim WHENEVER SUCCESSFUL;

Controlling Object Auditing

The command that starts auditing activity against specific database objects has the following syntax: AUDIT object_option[,...] [ALL] ON [[schema.]object_name] [DIRECTORY directory_name] [DEFAULT] [BY [SESSION] [ACCESS]] [WHENEVER [NOT] SUCCESSFUL] Object auditing isn't user-specific You can't select specific users when auditing on objects. All users performing an audited task on an object will cause audit records to be written, either a session record per connection or one record per executed statement.
q q

q q

AUDIT and ON are the required keywords for the object auditing command. object_option[,...] is an option or a comma-separated list of options from Table 11.3. You must include at least one object option if you don't use the keyword ALL. ALL includes all valid object auditing options. It must be included if you don't include any individual options. schema names the object owner. Your own schema is assumed if it's not included. object_name is the name of the object on which auditing is to be started, or a synonym for the object. You must include an object name if you don't include the DIRECTORY or the DEFAULT option. DIRECTORY directory_name identifies the name of an object directory to be audited. You must identify a directory if you don't include an object name or the DEFAULT option. DEFAULT indicates that you want the auditing options to be applied automatically to all new objects created in the schema. You must include the DEFAULT option if you don't include an object name or the DIRECTORY default.

http://www.informit.com/content/0789716534/element_011.shtml (7 of 18) [26.05.2000 16:49:11]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8
q

BY SESSION/ACCESS determines whether all use of an audited action is summarized into a single audit record per user session (BY SESSION), or if a separate audit record will be generated each time the action is performed (BY ACCESS). If you omit this option, audited actions will be summarized at the session level. WHENEVER [NOT] SUCCESSFUL determines whether only successful (SUCCESSFUL) or unsuccessful (NOT SUCCESSFUL) attempts to use the audited action are recorded. If you omit the WHENEVER option, all actions-successful and unsuccessful-are audited.

Different object auditing options are available, depending on the type of object you're auditing. Table 11.3 identifies which options can be selected for each object type. The optional keyword, ALL, will begin auditing every option that can be audited for the object type, as shown in Table 11.3. Table 11.3 Object auditing options Object Option: ALTER AUDIT COMMENT DELETE EXECUTE GRANT INDEX INSERT LOCK READ RENAME SELECT UPDATE Table: X X X X X X X X X X X View: Sequence: Procedure Snapshot: Library: Directory: FunctionPackage: X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X

X X X

Considerations for the DEFAULT audit option Unlike the other options, DEFAULT applies to the whole database, not just to the schema of the user issuing the command. Therefore, following the successful completion of an AUDIT... DEFAULT command, any object created by any user will begin to be audited if the object audit list in the command included one or more operations valid for that object type. Using the ALL option in an AUDIT...DEFAULT command will cause every possible audit action to be applied to every new object created subsequent to the command being executed. You can enable object auditing on only one object at a time with an AUDIT command. The object can be a schema object or an object directory in which BFILE objects are stored. Instead of either of these, you can also enable auditing by default via the DEFAULT option. Default auditing does nothing to any existing objects, but will apply the selected audit options to any new object for which the option is valid. Any auditing that commences as a result of default auditing can be stopped with the appropriate NOAUDIT command. However, the command has to be issued against each individual object. The NOAUDIT...DEFAULT command will prevent further objects from being audited by default, but won't terminate any auditing already being performed. Special privileges for the AUDIT...DEFAULT option Due to its far-reaching capabilities, you can't successfully issue the AUDIT...DEFAULT command unless you've been granted the AUDIT SYSTEM privilege. The AUDIT ANY system privilege isn't adequate. There's a version of the NOAUDIT command for terminating object auditing. Its syntax follows that of the AUDIT command, except that it doesn't include the BY ACCESS/BY SESSION option. Just as with the system privilege and statement option NOAUDIT command, you can stop the auditing of successful or unsuccessful actions on the object actions with the appropriate WHENEVER clause option, or end auditing altogether on the chosen action(s) by completely omitting the WHENEVER clause.

http://www.informit.com/content/0789716534/element_011.shtml (8 of 18) [26.05.2000 16:49:11]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

Reviewing Audit Records

The data dictionary contains two types of views related to auditing. The first type identifies which items are being audited; the second type, built on the AUD$ table, shows the audit records from various perspectives. Before you turn auditing on for your database with the AUDIT_TRAIL parameter, or anytime auditing is active, I recommend that you monitor the former type of view (listed in Table 11.4). This will help you identify any auditing that you don't think should occur, thus avoiding unexpected and unnecessary accumulation of audit records. Scripts to manage audit-related data dictionary views Oracle provides two SQL scripts to manage the data dictionary views associated with auditing: CATAUDIT.SQL and CATNOAUD.SQL. CATAUDIT.SQL creates the views and is run as part of CATALOG.SQL, which should be run following database creation. CATNOAUD.SQL drops the audit-related views should you not require them. This script won't drop the base auditing table, AUD$, owned by SYS. If you drop the views, you can recreate them later by running CATAUDIT.SQL again. Remember to run CATAUDIT.SQL and CATNOAUD.SQL while logged in as SYS (eith her directly or through the INTERNAL login). Table 11.4 Data dictionary views showing current audit options View Name: ALL_DEF_AUDIT_OPTS Description: Lists the object-auditing options that will be activated by default on all new objects for which they're valid auditing events. This view contains a single-row, one-entry-per-object option in the form of -/-, where either hyphen (-) can be substituted with an S or an A. The first hyphen, S, or A indicates whether the auditing will be performed on successful use of the command, and the second indicates that auditing will be performed on unsuccessful use. A hyphen means this option won't be audited, an S means that the auditing will be by session, and an A means that it will be by access. Lists the values from the ACTION# column from the AUD$ table and the related audit action type. Lists the object-auditing options on all objects. This view contains a single entry for each database object, whether it's being audited or not. The rows are structured like the one in the ALL_DEF_AUDIT_OPTS view, with a two-part value under each option (each part containing a hyphen, an S, or an A). Describes the system privileges being audited. There's one entry if the privilege is being audited systemwide, and a separate entry for each individual user being audited. Describes the statement options being audited. There's one entry if the privilege is being audited systemwide, and a separate entry for each individual user being audited. Lists the object auditing options on all objects owned by the user. This view contains a single entry for each object owned by the user, whether it's being audited or not. The rows are structured like those in the DBA_OBJ_AUDIT_OPTIONS view.

AUDIT_ACTIONS DBA_OBJ_AUDIT_OPTS

DBA_PRIV_AUDIT_OPTS

DBA_STMT_AUDIT_OPTS

USR_OBJ_AUDIT_OPTIONS

The second type of audit-related views provides formatted access to the audit trail. These views are generally more useful for examining the audit results than querying the AUD$ table directly. Of course, you can't use these views, or any SQL-based retrieval method, to examine audit records that you store in an external audit trail file. You must develop your own techniques to report on audit records created when you set the AUDIT_TRAIL parameter to the value OS. Table 11.5 describes the views built over the entries in the audit trail generated as the result of a selected audit action. The table shows the view name and the nature of auditing that creates the audit records displayed through the view. Table 11.5 Data dictionary views for querying audit records View Name: Audit Action(s) Reported:

http://www.informit.com/content/0789716534/element_011.shtml (9 of 18) [26.05.2000 16:49:11]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

DBA_AUDIT_EXISTS DBA_AUDIT_OBJECT DBA_AUDIT_SESSION DBA_AUDIT_STATEMENT DBA_AUDIT_TRAIL USER_AUDIT_OBJECT USER_AUDIT_SESSION USER_AUDIT_STATEMENT USER_AUDIT_TRAIL

All statements resulting in NOT EXISTS errors Any object audit option All connection attempts All uses of GRANT, REVOKE, AUDIT, NOAUDIT, and ALTER SYSTEM commands All audit trail entries All statements concerning objects All connections and disconnections for the user All uses of GRANT, REVOKE, AUDIT, NOAUDIT, and ALTER SYSTEM commands by the user All audit trail entries relevant to the user

Values for audited actions in the AUD$ RETURNCODE column For the views that show whether the audited action was successful, the RETURNCODE column will contain an Oracle message number: a 0 for success or an exception code from the error message if unsuccessful. SEE ALSO Descriptions of the SQL scripts used following database creation to complete the data dictionary table and view definitions,

Profiles and System Resources

To help you control the use of system resources by your database users, Oracle provides you with a special type of database object known as a profile. Every database is built with one profile, called DEFAULT, but you can add as many others as you need. Within each profile, you set limits on various system resources that can be consumed. When more of a resource has been used than the limit allows, Oracle will stop further processing and issue an appropriate error message. Every profile can define different limits for some or all controlled resources. You can choose to leave certain resources with no limitations on their use in any of your profiles. Each database user is assigned to one profile, and the user's database sessions can be constrained by the limits set in that profile. Profile resource limits won't be enforced, however, unless the database is running in a special mode. You can set this mode with a parameter in your initialization file or with an ALTER SYSTEM command. The parameter, RESOURCE_LIMIT, takes the values TRUE and FALSE; you should set it to TRUE if you want to enforce the resource limits assigned to your profiles. You should issue the following command if RESOURCE_LIMIT is set to FALSE but you want to start enforcing resource limits: Scope of the ALTER SYSTEM command The ALTER SYSTEM command won't change the value of the related parameter in your initialization file. If you enable or disable resource checking with this command, therefore, you should also modify your RESOURCE_LIMIT parameter to the desired value. This will ensure that the database will continue to behave as you want it to should the instance need to be restarted for any reason. ALTER SYSTEM SET RESOURCE_LIMIT = TRUE; You can, of course, issue the command with the FALSE option to turn off resource checking, whether it was begun with an earlier ALTER SYSTEM command or due to the parameter value being set to TRUE. The ALTER SYSTEM SET RESOURCE_LIMIT command doesn't change the behavior of any active user sessions when the command is issued, so users can still encounter limits on their work even if the command turns off resource checking with the FALSE option. If you don't need to restrict your users with resource limits, ensure that you run your database with the RESOURCE_LIMIT parameter set to FALSE. Even if your users are all assigned to a profile that allows them unlimited resource usage, Oracle must locate the related profile and store the limits in the user's session data every time a new session starts when your system is running with resource checking enabled. This overhead can be avoided, and if you discover the parameter is set to a wrong value you should temporarily change the database behavior until you can change the parameter and restart the instance. The following command will prevent resource checks from being initiated in any new sessions:

http://www.informit.com/content/0789716534/element_011.shtml (10 of 18) [26.05.2000 16:49:11]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

ALTER SYSTEM SET RESOURCE_LIMIT = FALSE;

Combined Resource Limits

You can limit the consumption of different system resources through individual settings in a profile or through a combined usage value. You can even include both limit types within a single profile. You should consider setting up a combined usage limit when your database is running on a system where users are being monitored for their consumption of resources with a standardized unit of measure. Typically, these units-known as "computer resource units," "computer billing units," or "computer service units"-are composed of various amounts of one or more of CPU consumption, memory usage, disk I/O, and connection time. You can use what Oracle calls composite limits for your profiles. These can reflect the same ratio of resource usage as your system's service units. With them, you can terminate a user's session after it consumes more than a predetermined number of such service units. Before you can take advantage of composite limits, you must define the components to be included and their respective weightings. You use the ALTER RESOURCE COST command to complete these definitions and you can query the data dictionary view RESOURCE_COST to see the current settings. The command allows you to set a weight for each of the following: q The amount of CPU time used by a session q The session's connect time q The number of Oracle blocks read during a session q The amount of memory dedicated to the session (private SGA) The ALTER RESOURCE COST command's full syntax is as follows: ALTER RESOURCE COST [CPU_PER_SESSION integer] [CONNECT_TIME integer] [LOGICAL_READS_PER_SESSION integer] [PRIVATE_SGA integer] integer is the weight assigned to a unit of the resource's consumption, as shown in Table 11.6. You must include at least one resource in the command and as many of the other resources as you need. Table 11.6 in the following section shows the optional keyword to use in the ALTER RESOURCE COST command for these options and the unit of measure for each one. If you don't include one of the keywords in the command, it will retain its current value (as shown in the RESOURCE_COSTS view). Each resource has an initial value of 0 when the database is created. A composite limit is reached when the sum of the resources, multiplied by their assigned weights, exceeds the value of the limit. For example, assume you issued the following command and a session has accumulated 10 seconds of CPU time and accessed 2,000 Oracle blocks while being connected for 30 minutes: ALTER RESOURCE COST CPU_PER_SESSION 50 CONNECT_TIME 10 LOGICAL_READS_PER_SESSION 100 PRIVATE_SGA 0; It will have consumed the number of service units (shown here) toward a composite limit: 50 * 1000 (1/100's seconds CPU time) + 10 * 30 (minutes connect time) + 100 * 2000 (Oracle blocks accessed) Calculating composite service units Using this example, CPU has a weight of 50 and a unit of measure of 1/100 second, so 10 seconds contribute 5010/(1/100) = 50,000 CPU per session. Connect time has a weight of 10 and a unit of measure of minutes, so 30 minutes contribute 1030 = 300. Logical reads have a weight of 100 each, so 2,000 blocks accessed contribute 1002,000 = 200,000. Private SGA has a weight of 0 per byte, so all bytes contribute nothing more to the overall total. The sum of these factors gives the resulting total composite service units: 50,000+300+200,000+0= 250,300.
http://www.informit.com/content/0789716534/element_011.shtml (11 of 18) [26.05.2000 16:49:11]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

Creating Profiles
You create a new profile by using the CREATE PROFILE command. This command uses different keywords to set limits for the different resources. Table 11.6 identifies these keywords and lists which resources, plus their units of measure, they control. The table also indicates which of these resources contributes to composite limits, as discussed in the preceding section. Table 11.6 Resource limits controlled by profiles Part of Composite System Resource: Concurrent sessions Session CPU CPU per call Session elapsed time Inactive session Keyword for CREATE PROFILE and ALTER RESOURCE COST Commands: SESSIONS_PER_USER CPU_PER_SESSION CPU_PER_CALL CONNECT_TIME IDLE_TIME Unit of Measure: Number 1/100 second 1/100 second Minutes Minutes X X X Part of Composite Limit: X

Oracle blocks accessed in session LOGICAL_READS_PER_SESSION Number Oracle blocks accessed per call LOGICAL_READS_PER_CALL Number Session memory Service units PRIVATE_SGA COMPOSITE_LIMIT Bytes Number

For those resources that can be limited by session and by call, the session value includes all work done since connecting to the database, and the call value includes only the work done during a single database call, such as a parse, execute, or fetch. SEE ALSO Information about the parse, execute, and fetch database calls, The syntax for the resource management options of the CREATE PROFILE command is as follows: CREATE PROFILE profile_name LIMIT resource_key_word [integer][K|M] [UNLIMITED] [DEFAULT] [...] CREATE PROFILE profile_name LIMIT are the keywords to create the profile with the given name. resource_key_word is a valid keyword from Table 11.6. integer is the value of the limit assigned to the resource; this must be included if neither the UNLIMITED nor DEFAULT options are included for the resource. K and M are optional abbreviations for kilobytes and megabytes, respectively, that can be used only with the PRIVATE_SGA keyword. UNLIMITED removes any limitation on the resource from the profile. This must be included if neither a resource limit value nor the DEFAULT option is included for the resource. DEFAULT indicates that this resource will be limited by the current value in the DEFAULT profile at the time the profile is invoked. This must be included if neither a resource limit value nor the UNLIMITED option is included for the resource. [...] indicates that two or more resources can be controlled in a single statement.

q q

Privileges to manage profiles Profiles aren't members of any schema, so any user with the necessary privileges can issue commands against profiles created by any other user. The system privileges to create profiles, to alter resource limits in existing profiles, and to drop profiles are CREATE PROFILE, ALTER PROFILE, and DROP PROFILE, respectively. You don't have to include a limit for every resource in a profile when you create it. Any unnamed resource will be treated as though it were assigned the DEFAULT keyword. In other words, any time the profile is used, the values for unassigned resource limits will be taken from the corresponding values in the DEFAULT profile.

http://www.informit.com/content/0789716534/element_011.shtml (12 of 18) [26.05.2000 16:49:11]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

Assigning Profiles
To enforce the limits in a profile you've created, you need to assign that profile to one or more users. You can assign a profile by using the PROFILE option of either CREATE USER or ALTER USER. The RESOURCE_LIMIT must also be activated for the profile limits to take any effect, of course, although the profile can be assigned at any time. A profile must always be part of a user definition If you don't want a userid to continue to be assigned to one of your own profiles, you must issue the ALTER USER command to reassign it to the default profile, DEFAULT. A userid must always be associated with an existing profile. When you change the profile assigned to a user with the ALTER USER command, the new profile will become effective the next time someone connects to the database with that userid. However, any currently active session will continue to function under the limits imposed by the profile assigned to the userid at the time they were initiated. When profile limits are in effect and a user exceeds a limit, the outcome will depend on the specific resource type: q When a session exceeds the session limit on CPU time, connect time, blocks accessed, memory used, or on the composite limit, the transaction being processed at the time will be rolled back and the user will be disconnected from the database. q When a statement exceeds a call limit, the current statement is rolled back but the session and the current transaction remain active. q When a session remains inactive for longer than the idle time limit, any current transaction will be rolled back and the session terminated the next time a statement is attempted. q The database connection is refused when a new session exceeds the number of concurrent sessions allowed by the profile. Inactive really means inactive to Oracle For the purposes of the idle session timeout, a session is considered inactive only if no statement is being processed by the server. Unlike some operating system connection time-outs, a long-running statement or query isn't considered inactive, even if the application isn't receiving or displaying any results for more than the idle time limit. SEE ALSO The complete syntax of the CREATE USER command, The complete syntax of the ALTER USER command,

Altering Profiles
You use the ALTER PROFILE command to change one, some, or all of the resource limits defined in a profile. Use exactly the same syntax as that for the CREATE PROFILE command, substituting only the word ALTER for the word CREATE. The only restriction on the ALTER PROFILE command is on the DEFAULT profile-you can't use the DEFAULT keyword when changing one of its resource limits. You can, however, change any of the resource limits in DEFAULT profile to any other valid value. Any resource not named in an ALTER PROFILE command will retain the current limit's value. Include the resource name with the keyword UNLIMITED to remove a limit from a specific resource. As when you assign a profile to a user, any current session settings aren't changed when you execute the ALTER PROFILE command. Only new sessions started by users assigned to the altered profile will be accorded its new resource limits. Watch for unexpected side effects when changing the DEFAULT profile The user SYS needs to retain the ability to use unlimited resources to complete essential database activities. These include the execution of recursive SQL statements and processing performed by the background processes. SYS uses the UNLIMITED resource limits (set by default in the DEFAULT profile) when resource limits are activated. Therefore, you shouldn't reduce any of these limits unless you've already built an alternate profile with unlimited resources and assigned it to SYS. You also might want to assign this alternate profile to the SYSTEM userid and your primary DBA user accounts.

http://www.informit.com/content/0789716534/element_011.shtml (13 of 18) [26.05.2000 16:49:11]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

Dropping Profiles
You can drop any profile except the DEFAULT profile by using the DROP PROFILE command: DROP PROFILE profile_name [CASCADE] The CASCADE option must be included if the profile is still assigned to one or more users. As with other changes to profiles, any sessions started with the profile assigned to them will continue to be limited by the profile limits even after it's dropped. The userids assigned to the dropped profile will be assigned to the DEFAULT profile automatically, and any new sessions started by these userids will be constrained by DEFAULT's limits.

Profiles and Password Management

Oracle8 allows you to control your database users' management of their passwords. The password-management features have many capabilities: q Pre-expire new passwords q Expire passwords after a specific period of time q Allow a grace period between password expiration and the deactivation of the userid q Prevent the reuse of a password for a specific period of time q Lock and unlock accounts manually q Force a password to meet certain complexity requirements The same profiles that you utilize to limit system resource use can be used to provide you with password-management tools. The profiles already contain entries for password-management functions, but these aren't enabled unless you run a special script. This script, UTLPWDMG.SQL, can be found in the same location as the other scripts discussed in Chapter 2 "Creating a Database." To activate the password-management features, execute this script after connecting to Oracle with the SYS userid. The script has two components: The first changes the password-related entries in the DEFAULT profile and the second builds a password-complexity function in the database. The script also causes the database to begin password-checking activities against all profiles. Password management is independent of resource management Although the password-management features are controlled by database profiles, they aren't enabled and disabled the same way the resource limits are controlled by profiles. Resource limit checks are activated by starting the database with the initialization parameter RESOURCE_LIMIT set, or by issuing ALTER SYSTEM to change RESOURCE_LIMIT to TRUE. Checking can be deactivated by changing the value to FALSE in the parameter file or with the ALTER SYSTEM command. Password management is activated by running the UTLPWDMG.SQL script and stays active from that point on. The RESOURCE_ LIMIT value has no impact on password-checking activities.

Creating Password-Management Profile Entries

You use the CREATE PROFILE and ALTER PROFILE commands to set the values in profiles that determine the password characteristics for the users who are assigned to them. The syntax for these commands, when addressing password options, is as follows: Representing portions of days for password options You can set values for password options that provide time limits measured in whole days, or in portions of a day. For example, two weeks would be represented by the number 14, whereas one hour could be represented as 1/24 or 0.0416667. The CREATE and ALTER PROFILE commands accept any of these formats for days, integers, fractions, and decimals. However, you can't mix whole numbers and fractions in a single expression. One and a half days would have to be entered as 1.5, not as 1fi. The smallest time increment you can use for a password option is one second, represented by 1/86400 or 0.000015741. CREATE|ALTER PROFILE profile_name LIMIT password_keyword [expression][UNLIMITED][DEFAULT][NULL] [...]

http://www.informit.com/content/0789716534/element_011.shtml (14 of 18) [26.05.2000 16:49:11]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8
q q

q q

q q

CREATE or ALTER will build a new profile or modify an existing one, respectively. PROFILE profile_name LIMIT are the keywords to identify the profile being processed by the command. password_keyword is a valid password option keyword from Table 11.7. expression contains a value valid for the password option, as indicated in Table 11.7. Expressions that represent the numbers of days can contain whole numbers, fractions, or decimal representations of days and partial days. UNLIMITED removes any limitation on the password from the profile. DEFAULT indicates that this password option will conform to the current setting in the DEFAULT profile at the time the profile is invoked. NULL indicates that no password complexity checking is to be performed. This option is valid only with the PASSWORD_VERIFY_FUNCTION password option. [...] indicates that two or more password options can be controlled in a single statement.

Only one value-expression, UNLIMITED, DEFAULT, or NULL-can be entered for any password option in a given statement. If either password option-PASSWORD_REUSE_TIME or PASSWORD_REUSE_MAX-is set to a numeric value, the other must be set to UNLIMITED. Mixing resource limits and pass-word options You can issue a single CREATE or ALTER PROFILE command that contains resource limit and password option values. Table 11.7 Password management options Function: Keyword for CREATE or ALTER Expression Values PROFILE Commands: for CREATE or ALTER PROFILE Commands: FAILED_LOGIN_ACCOUNTS An integer PASSWORD_LIFE_TIME PASSWORD_REUSE_TIME PASSWORD_REUSE_MAX An integer, decimal, or fractional number of days An integer, decimal, or fractional number of days An integer

Lock an account after a number of tries to log in Expire an unchanged password after a number of days Prevent reuse of a password for a number of days Prevent reuse of a password before some number of password changes Keep a password locked for a number of days after consecutive failed login tries Provide a number of days for warnings to be given before locking accounts with expired passwords Name a function that examines the password for desired characteristics

PASSWORD_LOCK_TIME

An integer, decimal, or fractional number of days An integer, decimal, or fractional numberof days

PASSWORD_GRACE_TIME

PASSWORD_VERIFY_FUNCTION Name of a password complexity function

The password-complexity checking (verify) function is explained in detail in the next section. When you run the UTLPWDMG.SQL script, you can see the password settings in the DBA_PROFILES data dictionary view. This view lists the profile options and the current values for each defined profile, including DEFAULT. The RESOURCE_TYPE column will have the value PASSWORD for the entries associated with password management, and the value KERNEL for the entries associated with operating system resource limits. The keywords DEFAULT or UNLIMITED will appear in the LIMIT column when a specific value isn't assigned. If the PASSWORD_VERIFY_FUNCTION option was entered as NULL, the value in the DBA_PROFILES view will be UNLIMITED in the LIMIT column. Considerations for password management

http://www.informit.com/content/0789716534/element_011.shtml (15 of 18) [26.05.2000 16:49:12]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

You may want to use different password-management rules for different categories of users, which you can easily accomplish by building different profiles for each group and assigning them to the appropriate userids. You should make sure that any users who remain assigned to the DEFAULT profile won't have inappropriate password management options defined by it. Pay particular attention to the special database userids, SYS and SYSTEM, and to other IDs used by your operations staff or associate DBAs, as well as any userids used for processing batch jobs. If the password options applied to DEFAULT by the UTLPWDMG.SQL script aren't appropriate for these users, you should either alter the DEFAULT profile or build and assign profiles to control their pass-words according to their particular needs. When assigning profiles to manage passwords, the profile options that are checked are those belonging to the user whose password is being assigned or changed, not those in the profile assigned to the user issuing the command. For example, if the userid SYSTEM has a profile with PASSWORD_REUSE_MAX set to UNLIMITED, a user connected as SYSTEM could issue the following command an infinite number of times without error: ALTER USER system IDENTIFIED BY manager; However, if the user SCOTT were assigned to a profile with PASSWORD_REUSE_MAX set to 1, a user logged into the SYSTEM userid with the profile as described couldn't issue the following command more than once successfully: ALTER USER scott IDENTIFIED BY tiger; The limit on password reuse set by Scott's profile takes effect, not the limit in the profile assigned to SYSTEM.

Checking for Password Complexity

If you want to force your users' passwords to comply with certain rules, such as a minimum number of characters or at least one digit in the password, you must use a PL/SQL function to perform the requisite checks. Oracle provides VERIFY_FUNCTION as the default function for checking password complexity. The function returns a Boolean value of TRUE if the password passes all the checks, or exits with a return code in the range ORA-20002 to ORA-20004 if the password fails a check. The function is created by the UTLPWDMG.SQL script, which also adds its name to the PASSWORD_VERIFY_FUNCTION option in the DEFAULT profile. If you want to create your own password-complexity function(s), you should use VERIFY_FUNCTION as a model. You can examine the function's code by querying the DBA_SOURCE data dictionary view or, more simply, by reading the UTLPWDMG.SQL script. (For more information about the PL/SQL language and the creation and management of PL/SQL functions, see Appendix A, "Essential PL/SQL: Understanding Stored Procedures, Triggers, and Packages.") You can activate your password functions as soon as they're created by naming them in the PASSWORD_VERIFY_FUNCTION option of the CREATE of ALTER PROFILE command. Users assigned to the profile will have any new passwords checked by your function. You may not want to continue to use the Oracle-supplied VERIFY_FUNCTION in the DEFAULT profile because it has some characteristics that you may not find appropriate. For example, the function would disallow the password tiger, which is used in a number of the demonstration scripts that create or use the SCOTT userid. It also disallows the password manager, which is the password expected in other scripts that Oracle requires you to run under the SYSTEM userid. You have three options for discontinuing the use of the default function by DEFAULT: q You can replace the VERIFY_FUNCTION with one of your own and change the DEFAULT profile to identify it: CREATE FUNCTION my_password_function (username VARCHAR2, password VARCHAR2, old_password VARCHAR2) RETURN boolean IS ... END; ALTER PROFILE default PASSWORD_VERIFY_FUNCTION my_password_function; You can modify the VERIFY_FUNCTION: CREATE OR REPLACE FUNCTION VERIFY_FUNCTION (username VARCHAR2,

http://www.informit.com/content/0789716534/element_011.shtml (16 of 18) [26.05.2000 16:49:12]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

password VARCHAR2, old_password VARCHAR2) RETURN boolean IS ... END; You can turn off complexity checking for the profile: ALTER PROFILE default PASSWORD_VERIFY_FUNCTION NULL;

To test the code in a password-complexity function, I recommend that you build and use the following SQL*Plus script: DECLARE status BOOLEAN; BEGIN status := &function_name (user, &new_password, &old_password); END; / Testing your own password-complexity functions If you decide to code your own PL/SQL functions to control password structure and complexity, you should develop a structure to test them. I recommend creating a userid and a profile for developing and testing the code. Also use a script file to hold the PL/SQL CREATE OR REPLACE FUNCTION command. Use the same name for each function while it's in development so that you don't continually have to modify the profile's PASSWORD_ VERIFY_FUNCTION entry. When you're sure that the function is working as you want it, you can copy the test script to a permanent storage location, alter the test name to a production name, and execute the script under the SYS userid. Be particularly careful when building your own password functions to be used with the DEFAULT profile or other profiles assigned to key userids such as SYS, SYSTEM, or your DBA accounts. You can then execute this script from a SQL*Plus session by providing the name of the function to be checked and testing values for new and old passwords when prompted. You can replace the USER function with the &USER substitution variable if you also want to test the function against different userids. Note that testing the function with this script won't cause any password changes to be stored in the database. Before completing this discussion of the complexity function, we should briefly discuss the Oracle-supplied VERIFY_FUNCTION function because, if you look at this code, two of the checks can cause some confusion: q Consider the check to guarantee that there's at least one each of the three types of character (letter, digit, and punctuation) somewhere in the password. The code uses string variables to hold the valid characters for each type. The string for checking punctuation contains the standard punctuation marks such as the period (.), comma (,), asterisk (*), and so on. This might lead you to believe that users can now create passwords with such characters embedded in them. You might also assume that any of the three character types can be used anywhere in a password. However, Oracle8 passwords must conform to the standard naming convention for Oracle objects. In other words, they must begin with a letter and contain only letters, digits, underscores (_), dollar signs ($), and pound signs (#). In order to use any other format or any other punctuation character, your users would have to enclose their passwords in double quotation marks (") each time they used them. q The other misleading check compares the old and new passwords. The intent of this check is to ensure that they vary by a certain number of characters. You can execute VERIFY_FUNCTION directly, supplying values for the username and the new and old passwords, to confirm that this check does indeed work as documented in the script. However, if you look closely at the code, you'll see that this particular check isn't performed if the input value for the old password is a zero-length string. Unfortunately, when the function is executed as part of a password-change command (ALTER USER user_name IDENTIFIED BY password), Oracle doesn't supply the old password because it doesn't know it. Passwords aren't stored in the database directly, but via a one-way encryption algorithm, which means that the current password can't be extracted from its encrypted version. The value for the old password is therefore always sent to the function as a blank string. The end result, of course, is that the function can't prevent the reuse of the same or similar password. You can overcome part of this limitation by using the PASSWORD_REUSE_MAX option. This can prevent the same password from being used twice in a row, or even from being reused until some defined number of different, intervening passwords have been used. Currently, there's no way to prevent similar passwords from being used right after each other. In addition, you can't code any of your own routines that depend on the value of the old password and have them work outside the test environment.
http://www.informit.com/content/0789716534/element_011.shtml (17 of 18) [26.05.2000 16:49:12]

informit.com -- Your Brain is Hungry. InformIT - Auditing Database Use and Controlling Resources and Passwords From: Using Oracle8

< Back

Contents

Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_011.shtml (18 of 18) [26.05.2000 16:49:12]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Selecting and Implementing a Backup Strategy

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q

Selecting a Backup Strategy Consider Using Recovery Manager

r

Recovery Manager Scripts Taking an Offline (Cold) Backup of a Tablespace Making a Full Cold Backup What Is a Block Split? Hot Backup Script Maintaining a Standby Database Activating a Standby Database Building a Failover Database with Replication

Performing an Offline (Cold) Backup

r r

Performing an Online (Hot) Backup

r r

Creating a Standby Database

r r

Using Replication as a Failover

r

Selecting a Backup Strategy

Oracle offers very rich and comprehensive backup and recovery options to suit every application's demand. The key is to implement a suitable strategy to meet the desired service levels between end users and the database operations group. A backup strategy suitable for a database depends on several factors, based on the nature of the application and the availability requirements. These factors will also decide whether your database is running in ARCHIVE LOG mode or in NOARCHIVE LOG mode. Table 13.1 lists the available database options that depend on the database's ARCHIVELOG mode. The backup option in NOARCHIVELOG mode Full cold backup is the only backup option available when the database is in NOARCHIVELOG mode. Table 13.1 Database mode and the available backup options Backup Type: Full cold backup Partial cold backup ARCHIVELOG Yes Yes NOARCHIVELOG Yes No

http://www.informit.com/content/0789716534/element_013.shtml (1 of 11) [26.05.2000 16:49:33]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

Hot backup

Yes

No

Recovery Manager and OS-level backup rely on making a copy of the users' data on a backup media-most commonly tape and disk. These methods preserve the data in a safe, inactive form. The data stored by these methods isn't used unless there's a loss of the current data for some reason. You can also copy the data stored in an Oracle database by using any of the following methods: Recovery Manager Recovery Manager is Oracle's new backup and recovery tool. Chapter 15, "Using Recovery Manager for Backup and Recovery," gives the details.
q

Export. Export enables you to make a logical backup of the data. You can export the full database, specific users, or specific tables. It provides a copy of the data stored in Oracle's proprietary format independent of the OS. Incremental backups are supported but aren't incremental in a strict sense, as any table is fully exported even if only one row has been modified. Export is used generally to transfer data from one database to another. Consider using Export for preserving contents of small lookup tables. However, the use of Export as a backup method is very limited in large databases due to its slower performance. Hot standby database. A hot standby database is normally used as part of the disaster recovery plan in mission-critical environments. A hot standby database needs machine resources equivalent to the primary database. Data can be made available to users at very short notice in a hot standby database, even if no other resource from the primary database is available. A hot standby database is used only when the cost of downtime for the database is very high and the downtime cost justifies the cost to devote redundant machine resources required to build the standby database. Replication. With this technique, you can maintain object-level replicas of user data. The data is asynchronously available to all users at all locations after a predefined replication interval. Use this method when the business requires that the same data be available from multiple databases. It uses a trigger-based technique to propagate changes among the database and is a resource-intensive process. "Using Replication as a Failover," later in this chapter, provides more detail on this subject.

In summary, all these methods offer unique advantages and need to be used depending on your business requirements. However, they can also be used to complement each other.

Consider Using Recovery Manager

Oracle8's Recovery Manager is an integrated backup and recovery tool. Consider the following salient features before deciding your backup strategy: Media management libraries The Media Management library contains the vendor-supplied software programs that supply the backup Media Manager. If you don't have any backup media installed on the machine, disks can be used to take back-ups.
q

Recovery Manager is integrated with Oracle. It performs backup operations with spawned Oracle processes. These processes use Media Management libraries, which interface with backup media. You don't have to specify an OS utility such as tar, Backup, dd, and so on to copy files. Recovery Manager greatly simplifies administrative tasks associated with backup and recovery. Several tasks-defining backup configurations, keeping a log of backup and recovery operations, automatically parallelizing the backup and restore with the resources available-can be easily automated with Recovery Manager. Recovery Manager detects any Oracle block splits and rereads these blocks to get a consistent view. Therefore, it isn't necessary to keep the tablespace in backup mode while performing a backup. However, you may perform a consistent cold backup of the database by using Recovery Manager. (See "What Is a Block Split?" later in this chapter for more information on block splits.) Recovery Manager doesn't back up Oracle blocks that have never been used, thus saving considerable backup time and space. Recovery Manager stores backed-up data in an Oracle internal format that can't be read by any other utility. Therefore, files backed up by Recovery Manager can be restored and recovered only with Recovery Manager.

If you aren't using Recovery Manager to perform backup and recovery, you can use OS-level physical backups for making copies of the data to protect against loss. This has so far been the most widely used method for
http://www.informit.com/content/0789716534/element_013.shtml (2 of 11) [26.05.2000 16:49:33]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

backup operations. There are important considerations to make while implementing a backup strategy: q You must run the database in ARCHIVE LOG mode in an online transaction system, where each and every transaction is important and it's necessary to recover up to the last committed transaction. q When the database is running in ARCHIVELOG mode, you can perform a full or partial cold or hot backup. q A full cold backup is the simplest backup to implement and should be the preferred method, unless availability requirements don't allow enough downtime to perform a full cold backup. A full cold backup can also be integrated with other non-Oracle files at the OS level. q Recovery Manager is a very flexible and powerful backup and recovery tool. Consider using the Recovery Manager for backup and recovery operations. Refer to Chapter 15 for more details on using the Recovery Manager. q Recovery Manager enables you to take true incremental backup of the database. You can back up only the modified blocks since the last backup was performed at the same or lower level.

Recovery Manager Scripts

Recovery Manager is an easy-to-use tool available in command-line interface (CLI) and graphical user interface (GUI) mode. (The GUI mode of the Recovery Manager is available through Enterprise Manager on the Windows NT platform.) Oracle also provides comprehensive scripts to start with. The scripts case1.rcv and case4.rcv provide comprehensive information about using the Recovery Manager to perform backup. These scripts are located in the $ORACLE_HOME/rdbms/demo directory on a UNIX system and the \ORANT\rdbms\demo folder on a Windows NT system. case1.rcv This script contains code for the following backup, restore, and recovery operations using Recovery Manager: Print the script case1.rcv while reading this part case1.rcv is a long script that shows how to write Recovery Manager scripts. Keep a printout of it handy while reading this section.
q q q q q q q q q q q q q q q q q q

Archive the current log Specify maximum corruption tolerance Mount the database Back up archived logs, back up the control file, and back up the control file with the copy command Back up system tablespace Full, whole database backup Incremental level 0, whole database backup Incremental level 1, whole database backup Incremental level 2, whole database backup Back up read-only tablespace Script to verify backup set isn't corrupt Restore and recovery Restore control file Restore all data files Recover all data files Restore and recover data files and control file Restore and recover data files Restore and recover a single tablespace (database open)

Keep the following in mind when you perform these functions: q Use the setlimit, filesperset clause to enforce the following restrictions: No single backup piece is greater than 2GB. Puts a maximum of 20 archived logs in any one backup set.

http://www.informit.com/content/0789716534/element_013.shtml (3 of 11) [26.05.2000 16:49:34]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

No more than 200 buffers are allowed to be read per file per second, which limits the effect the backup file scan has on online users and batch jobs.
q q q

A channel can have a maximum of only 32 files open at any one time. Includes, at the most, six files in one backup set.

Continue even if any of the data files is inaccessible, leaving unavailable data files. Skip data files that belong to read-only tablespaces and are offline. Back up the tablespace that has just been made read-only; it should be backed up once before being omitted from future backups by the skip readonly clause.

Making backups on disk with Recovery Manager To make backup sets to disk with Recovery Manager, change the channel allocation commands to be type disk and modify the format clause to include the full path name the backup is to be written to. Otherwise, the backup files will typically be written to the dbs directory in Oracle home. case4.rcv This script has the code for taking a consistent backup of a database using the Recovery Manager. It performs the backup operations with the following considerations: q Backs up database by using up to two tape drives. q The Media Manager has a maximum supported file size limit of 2GB, so any backup piece should be no larger than that size. q Each backup set should include a maximum of five files. q Includes offline data files and read-only tablespaces in the backup. q Terminates the backup if any files aren't accessible. q Opens the database after the backup completes.

Performing an Offline (Cold) Backup

Cold backup or consistent backup of a file is the backup taken when the file is offline (not being accessed by the database instance, for instance). Thus, a file's cold backup is its image copy as it existed when the file was taken offline. You can take a partial or full cold backup off the database. The following factors determine whether a full or a partial cold backup is taken: Enabling ARCHIVELOG mode By default, the database is created in the NOARCHIVELOG mode. Refer to Chapter 12, "Understanding Oracle8 Backup Options," for steps required to alter the database to ARCHIVELOG mode.
q

A partial cold backup can't be taken if the database is in NOARCHIVE LOG mode; you can't take the files offline for cold backup. Size of the database, speed of the backup media, and the time allowed for the database to be unavailable. If the database is in ARCHIVE LOG mode, you can offline tablespaces to back them up while the rest of the database is available to the users.

Taking an Offline (Cold) Backup of a Tablespace

To make a cold backup of a tablespace (other than system tablespace data files), take the corresponding tablespace offline by issuing the following command from Server Manager: You can't take the SYSTEM tablespace offline To make an offline tablespace backup, the database needs to be shut down. You can't perform a cold backup of the SYSTEM tablespace while the database is running because Oracle requires that it always be available. SQL> alter tablespace tablespace name offline ; Use any of the OS utilities tar, dd, and cpio on UNIX and Backup on Windows NT to copy the corresponding data files. When the backup is completed, use the following command to put the tablespace back

http://www.informit.com/content/0789716534/element_013.shtml (4 of 11) [26.05.2000 16:49:34]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

online: SQL> alter tablespace tablespace_name online ; Listing 13.1 lists a sample script that performs an offline backup of a tablespace on a UNIX system. This script does the following: Download this code You can download the TS_BACK-UP_ OFF.SH script from the Web at http://www.mcp.com/ info. You'll be asked to enter an ISBN; enter 0789716534 and then click the Search button to go to the Book Info page for Using Oracle8.
q q q q

Lists the data files belonging to the tablespace (lines 9-23) Takes the tablespace offline (lines 28-30) Backs up the files with the UNIX tar command (lines 38-49) Puts the tablespace back online (lines 51-56)

Listing 13.1 TS_BACKUP_OFF.SH-Performing an offline backup of a tablespace View Code Contents of offline tablespaces aren't available to users When a tablespace is made offline, user objects residing in it aren't available for use until it's put back online. Applications trying to use these objects will signal errors while the tablespace remains offline. Taking a tablespace offline You can't take offline a table-space that contains an active rollback segment. First take the rollback segment offline, then the tablespace.

Making a Full Cold Backup

To perform a full offline database backup, the database must be shut down. Listing 13.2 shows a script that performs a cold backup of a database on a UNIX system. This script performs the following sequence of operations: q Generates a list of all the database's data files, control files, and online redo log files (lines 36-53) q Shuts down the database (lines 59-62) q Backs up the default parameter file (line 68) q Backs up the files with the UNIX tar command (lines 72-76) q Starts the database after the backup (lines 81-84) Listing 13.2 COLD_BACKUP.SH-Performing a full cold backup of a database View Code Shut down all instances in an OPS environment In order to perform a cold backup of a database in an Oracle Parallel Server environment, you must shut down all instances accessing the database before copying the data files at the OS level. Shutdown immediate disconnects all active users Shutdown immediate disconnects users and rolls back all current uncommitted transactions. The data in uncommitted transactions is lost. If you don't want to forcefully disconnect the users for performing backup, use shutdown normal. Pay attention to the following points when performing a cold backup: q Use Shutdown immediate or Shutdown normal. Before copying the files, ensure that the database has been shut down properly by looking at the messages in the alertsid.log file.

http://www.informit.com/content/0789716534/element_013.shtml (5 of 11) [26.05.2000 16:49:34]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

Don't use Shutdown abort before a cold backup If you shut down the database with the ABORT option, you should restart the database in RESTRICT mode and use Shutdown normal before copying the database files.
q

q q

The sample script shown in Listing 13.2 backs up online redo log files. Be careful while restoring them, however; if you plan to perform recovery by using the archived redo logs after the restore, don't restore the online redo logs from the backup. If you do so, they will overwrite the current online redo logs on the system and you won't be able to perform a complete recovery. Always back up archived redo logs at regular intervals. Use an automated method to get a list of the files that are part of the database similar to the one used in the sample script cold_backup.sh in Listing 13.2. This will minimize the administrative work and human errors. If you decide to list the files manually for backup purposes, remember to modify the script after you add and delete data files, control files, and online redo log files to and from the database. An automated script that picks up the names of the data files from the database doesn't depend on the underlying database file architecture (such as OFA, Optimal Flexible Architecture) to discover the name of the files to back up. Thus, it doesn't restrict you to placing the database files in a predefined manner. If the database contains READ ONLY tablespaces, it's not necessary to back them up each time during a cold backup. A one-time backup of these tablespaces is necessary, however, after making them READ ONLY. You don't need to back up tablespaces that don't contain any permanent objects. Thus, you may not back up tablespaces that contain only rollback segments and are used as temporary tablespace. I recommend, however, that you include them in the cold backup for ease of operation during the recovery. If you have to perform recovery after a full restore and haven't backed up the temporary tablespace and the tablespace containing rollback segments, you might have to perform extra steps before completing the recovery. Always look at the log generated during execution of the backup script to ensure that the backup process completed properly. The sample script in Listing 13.2 is provided for a quick start; it doesn't include a comprehensive error-checking mechanism. You might want to enhance it for checking the status of the database prescript, the status of the backup device, the space available at the backup destination, the errors during backup, and so on. The initialization parameter file is an ASCII file and can be created with any editor. It isn't necessary to back this up with every cold backup. I recommend backing it up, however, as it involves resources of little significance and can save valuable time in case you lose it.

Performing an Online (Hot) Backup

A hot or online backup refers to the backup of a database's data files taken while the database is still open. Use online backup when business requirements don't permit the database to be unavailable for an interval long enough to perform the offline data files backup. To perform online backup of a data file, you need to put the corresponding tablespace in backup mode by using the following command: SQL>alter tablespace tablespace_name begin backup; Copy the file at the OS level after issuing this command. When the copy operation is completed, the tablespace should be returned to normal mode with the following command: SQL>alter tablespace tablespace_name end backup; Oracle copies entire block images of the modified data blocks on the first change to the blocks; this is done while the tablespace is in the backup mode. This increases the size of the redo log generated while the tablespace is in the backup mode. Follow these guidelines to minimize its impact on the performance: q Perform the online backup of a tablespace while its DML activity is minimal. q Don't keep the tablespace in backup mode longer than what's required to copy the appropriate data files. Start copying the files immediately after issuing the begin backup command, and issue the corresponding end backup command as soon as copying is done. View V$BACKUP Use the dynamic performance view V$BACKUP to determine the back-up status of the data files at any time. The ACTIVE value in this column indicates that the file is marked as now being backed up.
q

Ensure that a tablespace isn't left in backup mode for a very long period after a failure in backup process

http://www.informit.com/content/0789716534/element_013.shtml (6 of 11) [26.05.2000 16:49:34]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

during the middle of the backup.

What Is a Block Split?

An Oracle data block consists of multiple OS disk blocks. The Oracle DBWR process writes modified blocks to the disk through OS write calls, which write to the disk in chunks that are multiples of OS blocks. The same Oracle block may be read and written to the storage media at the same time. This simultaneous read/write operation causes what's known as an "Oracle block split." Figure 13.1 shows three Oracle blocks. Each Oracle block consists of four OS blocks-a total of 12 OS blocks. Assume that time t1, t2, and t3 are in increasing order, and consider the following situation: Figure 13.1 : Oracle blocks splits/fractures as DBWR writes the block while it's being copied at the OS level.
q q q

The backup process writes the first six OS blocks to the backup file at time t1. The DBWR process writes Oracle Block 2 at time t2. The backup process writes the next six OS blocks to the backup file at time t3.

Clearly, the backup file contains the first half of Oracle Block 2 before the update, and the second half after the update. Thus, an inconsistent copy of the Oracle block is present in the backup file. To handle this problem, Oracle writes the complete image of the modified blocks in its redo log while the tablespace is in backup mode. During the recovery process before applying the changes to a block, Oracle copies the block image from its redo log to the disk-making the block consistent-and applies the remaining redo for the block.

Hot Backup Script

You can use the HOT_BACKUP.SH script shown in Listing 13.3 to perform a hot backup of a database on a UNIX system. This script backs up all data files in online mode. It also appends a backup copy of the control file and all archived redo logs present in the archive destination directory. Tailor it to suit your environment. This script calls another script, bkp_ts.sh (shown in Listing 13.4), which issues the necessary begin and end backup commands and copies all the data files belonging to a tablespace. Listing 13.3 HOT_BACKUP.SH-Performing a hot backup of a database View Code Log switch after a hot backup Always back up the redo logs generated during a hot backup by forcing a log switch and then copying the generated archived redo logs. Listing 13.4 BKP_TS.SH-Performing a hot backup of a tablespace View Code

Creating a Standby Database

If a database is used for mission-critical applications, you want to avoid downtime at any cost. Modern hardware and software techniques enable you to increase the availability of systems approaching 100 percent. If for any reason the primary production database can't be made available to the end users, however, maintaining a standby may fulfill a business's availability requirements. The standby database can't be queried A standby database is constantly in recovery mode and not available for querying the data unless you activate it. Once a standby database is activated, it's no longer available as a standby unless you recreate it as another standby database. A standby database is a copy of the primary database that can be brought online with minimum delay when needed. This feature is part of Oracle's disaster recovery procedures. With this feature, you can maintain a hot standby copy of the production database, preferably at another remote site, which can be activated on short notice. When it's activated, this database will serve as the production database. You might need to start from the beginning to create a standby database for this database.

http://www.informit.com/content/0789716534/element_013.shtml (7 of 11) [26.05.2000 16:49:34]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

Create a standby database 1. Perform an online or offline backup of the production database using the proper procedures. (If your system is mission-critical, however, you would most likely perform an online backup.) It's good practice to perform an online backup while setting up the standby database for the first time; this will give invaluable experience in recreating the standby database whenever the standby database is put into production use. 2. Create a control file for the standby database by using the following command: SQL>alter database create standby controlfile as control_file_name; The following is a sample command session executing this command. Notice that the filename is included in quotation marks. Because the full path name isn't given for the file, the standby control file is created in the default location-the $ORACLE_HOME/dbs directory. SVRMGR> alter database create standby 2> controlfile as 'control.std'; Statement processed. The standby control file Standby control file contents are different from those in the original control file. You should not use a backup copy of the original control file generated by the backup control file command instead of this file. 3. Archive the production database's current online log files with this command: SQL> alter system archive log current ; This command forces a log switch and then archives the current redo log file group. The database must be open to issue this command. This command may not be required if you've performed an offline backup of the database in Step 1, as data files are checkpointed (synchronized) before an offline backup. If the database is open, issuing this command is important because it ensures consistency among the data files in Step 1 and the control file in Step 2. 4. Transfer all files generated in Steps 1 through 3 to the system where you want to build the standby database. The standby should be built on a similar system The standby database must be built on the same hardware platform as that of the primary database. It's recommended there be similar architecture and software on both machines. Create the initialization parameter file for the standby database. It's highly desirable and also recommended to have the standby database's parameters similar to the primary database parameters, because the standby database will be used as the primary database after the failover. Keeping most parameters the same will help you avoid any surprises during its operation. Table 13.2 lists important initialization parameters related to the standby database configuration. Table 13.2 Parameters related to standby database Parameter Name: COMPATIBLE DB_FILES Description: This parameter must be the same on the primary and standby databases. This parameter and the MAXDATAFILES clause of the CREATE DATABASE or CREATE CONTROLFILE commands specifies the maximum number of data files for the database. Keep it the same, as the number of data files allowed/needed will be identical at both places. This parameter specifies the name of the control files used for the database. Both databases should point to different files. Their names can be the same, as they're located on different machines. This parameter is set only on the standby database. Use this parameter only if the directory paths to the data files are different at both sites.

CONTROL_FILES

DB_FILE_NAME_CONVERT

http://www.informit.com/content/0789716534/element_013.shtml (8 of 11) [26.05.2000 16:49:34]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

LOG_FILE_NAME_CONVERT

This parameter is the same as DB_FILE_NAME_ CONVERT, except that it applies to the online redo log members.

Maintaining a Standby Database

After transferring all data from the primary database to the host machine for the standby database, you need to keep the standby database closely synchronized with the primary database. Keep the standby synchronized with the primary database 1. Start the standby database instance in the nomount state with the following command: svrmgrl> startup nomount ; Automate the archive log transfer to the standby database Archived redo logs generated at the primary site need to be applied to the hot standby on a regular basis. In a production environment, it's desirable to automate the process of transferring and applying the archived redo log to the standby database. Use the operating system's or Oracle's job-scheduling features to implement this. 2. Mount the standby database in the standby exclusive mode by using the following command: svrmgrl> alter database mount standby database; 3. Prepare the standby database for recovery to apply the archived redo logs received from the primary database: svrmgrl> recover [from location] standby database until cancel; 4. Oracle will prompt the next log to be applied. Apply all the logs that have been generated so far and then cancel the recovery. If the archived redo logs to be applied are being made available in the directory specified by LOG_ARCHIVE_DEST, the FROM location clause doesn't need to be specified. Because it's a recurring process, you might want to automate it by using Oracle job queues or operating system job-scheduling features. Using the Unrecoverable Option on the Primary Database Unrecoverable data loads and the no-logging feature associated with tables and indexes that don't generate redo logs aren't propagated to the standby database. If you activate the standby database without paying attention to them, the affected target objects will be marked as corrupted. You can drop the affected tablespace at the standby database after the activation, or recreate it by using a fresh backup of the concerned tablespace at the primary site and transferring it over to the standby database before activation. Adding and Deleting Data Files If you add a data file at the primary database, you should make a copy of the added data file and add it to the standby database before applying the redo log that was generated while the file was added. This adds the new file to the standby control file and will allow the standby database's recovery to proceed as normal. A dropped data file at the primary site is also removed from the standby database control file when the corresponding redo log is applied. Adding, Removing, and Clearing Redo Log Files The standby database's online redo log files Online redo log files for a standby database are initialized when the standby database is activated. You might consider using the alter database clear logfile command to initialize the online redo logs in advance; this saves time while activating the standby database. Adding and removing online redo logs doesn't have any impact on the configuration of the standby database's redo log con-figuration. Performing an incomplete recovery at the primary database and opening it with the resetlogs option will invalidate the standby database, as the resetlogs option initializes all the online redo logs and resets the log sequence number at the primary site.

http://www.informit.com/content/0789716534/element_013.shtml (9 of 11) [26.05.2000 16:49:34]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

Activating a Standby Database

The standby database is mounted and is in recovery mode. Issue the following command to activate it and make it available to the users: svrmgr> alter database activate standby database; Standby databases become normal databases after activation As soon as you activate the standby database, it no longer remains a standby database. It becomes a normal database totally independent of its primary database. If needed, use it to recreate another standby data-base. Be careful to apply all the archived and unarchived redo logs from the primary database before activating the standby database. Once you activate the standby database, all the transactions from the primary database that are present in the unapplied redo logs are lost.

Using Replication as a Failover

Oracle offers the following solutions for implementing highly available systems: q Oracle Parallel Server q Hot standby database (discussed earlier in this chapter) q Oracle Failsafe option q Replication For information on the Oracle Failsafe option, available on Windows NT cluster systems, refer to the Oracle Fail Safe Concepts and Administration Guide from Oracle Corporation. SEE ALSO An overview of Oracle Parallel Server options, Oracle's Replication option enables you to maintain the same data in multiple databases. By using replication, you can design applications that are available to end users all the time. If one database fails for any reason-including the unavailability of its host site due to natural catastrophe-the sites having the replicated database can be used by the application. If you're maintaining data at more than two locations, the application can access the data as long as one replicated site is functioning. The nice thing about this scheme is that it can be completely transparent to end users. If the database instance they were working on goes down, the application can connect the users to another replicated site immediately without the users knowing that. Look at this option's important features with respect to a fail-over strategy: q The Replication option lets you choose the objects to be replicated at the table level. Thus, if only part of the data is critical for the business application, only that data can be replicated. q There are no restrictions on the hardware and software architecture of the machines hosting the database's participating in a replicated environment. One database could be on a UNIX machine, another could be on a Windows NT box, and so on. You can replicate databases among platforms as long as the Oracle versions residing on them are compatible with each other. q The systems having replicated databases need not be clustered or part of an MPP (massively parallel platform) system as required while implementing the Oracle Parallel Server or Failsafe option. Common network communication is used for data transfers among the machines having a replicated database. q Due to the hardware requirements, the machines used for OPS and Failsafe are located in physical proximity; therefore, these two options don't provide protection against site failure. The Oracle Replication option and the hot standby database can have the remote database at different geographical locations and provide protection against site failure. q The Replication option involves significant overhead in terms of network and other system resources; as a result, it's generally not suitable for a very high volume transaction system. q Data is propagated to other sites asynchronously. This implies that when a database participating in an replicated environment fails, some transactions committed in it may not be propagated at the other site until the failed database comes back up. Replication shouldn't be used for failover if the application can't tolerate this type of inconsistency. SEE ALSO Information on advanced replication options,

http://www.informit.com/content/0789716534/element_013.shtml (10 of 11) [26.05.2000 16:49:34]

informit.com -- Your Brain is Hungry. InformIT - Selecting and Implementing a Backup Strategy From: Using Oracle8

Building a Failover Database with Replication

To use replication as a means providing increased availability, you can choose from the following options: q The read-only snapshots should be used if the backup site is to be used to provide read-only access to the data. In this situation, DML operations can't be performed until the primary site comes back up. q Provide full access to the data by using multimaster replication. The data is available for read as well as for DML access from all the sites all the time. Setting Up Read-Only Failover Database with Snapshots Set up snapshots to provide read-only access to data with the master site 1. To set up snapshot replication, you need to decide which tables need to be replicated. If the application doesn't require all the tables at the failover site, you can save resources by not configuring snapshots for tables that aren't needed. Refer to Oracle8's Server Replication Manual for complete details Detailed instructions on setting up snapshots and replication is beyond the scope of this book. An overview of the tasks to be performed is provided; please refer to the Oracle8 Server Replication manual for more details and command syntax. 2. Create the snapshot schema at the snapshot site and then establish database links to the master site. 3. Create a snapshot log at the master site to allow fast refresh of the snapshots. If you don't create a snapshot log, every refresh will be a complete refresh (which may not be acceptable in most cases). 4. Create a snapshot at the snapshot site. During this process, all the data from the master site is propagated to the snapshot site. 5. Create snapshot groups and assign snapshots to them. This is the process where you define the snapshot update interval. Setting Up Failover Database with Multimaster Replication The steps required to set up multimaster replication are very similar to setting up snapshots. Set up multimaster replication 1. Set up the replication environment at the master definition site. This site includes creating the replication owner and granting necessary privileges to it, defining a database link to be used for replication, and scheduling jobs to replicate changes from one site to another at regular intervals. 2. Decide which tables need to be replicated and who owns them at the master definition site. Create the schema and the tables to be replicated at the master definition site. 3. Repeat the functions from Step 1 on the remote (master) site. 4. Repeat the functions from Step 2 on the remote (master) site. 5. Create a master replication group at the master definition site. 6. Add replication objects to the replication group and generate the replication support for these objects. 7. Start the replication process by enabling changes in the replication group from QUIESCED mode to NORMAL mode. After the replication is set up, Oracle maintains the data at all participating sites. No additional steps are required to maintain the replicated data, except to resolve error conditions that might result from what's called the CONFLICTS. Refer to the Oracle8 Server Replication Manual for more details. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_013.shtml (11 of 11) [26.05.2000 16:49:34]

informit.com -- Your Brain is Hungry. InformIT - Using Oracle8 - Database - Oracle

You are here : Home : Using Oracle8 Featured Author

Using Oracle8
Exact Phrase All Words Search Tips

Megh Thakkar Megh Thakkar is the Director of Database Technologies at Quest Software in Australia. Previously, he worked as a technical specialist at Oracle Corporation. Megh is the author of E-Commerce Applications Using Java and Oracle 8I From Scratch (Que, 2000).

Author: David Austin Web Price: $29.99 US Publisher: Que ISBN: 0789716534 Publication Date: 7/22/98 Pages: 832 Table of Contents Save to MyInformIT Using Oracle8 is the ultimate DBA reference book for the everyday Oracle database administrator. This book helps DBAs administer, troubleshoot, tune, backup, and upgrade your systems--the essential tasks that every DBA faces. Using Oracle8 does not provide history, strategies, theories, and other nonessential information to dilute the text. You are be able to turn to a particular topic and find the essential information you are seeking. A concise writing style and the power-packed SideNotes and cross-references make this an extensive Oracle book--providing extensive use of second color in cross-referencing and indexing. Table of Contents Introduction 1 -Introducing Relational Databases and Oracle8 2 -Creating a Database 3 -Migrating an Oracle7 Database to Oracle8 4 -Managing with Oracle Enterprise Manager (OEM) 5 -Managing Your Database Space 6 -Managing Redo Logs, Rollback Segments, and Temporary Segments 7 -Adding Segments for Tables 8 -Adding Segments for Different Types of Indexes 9 -Creating and Managing User Accounts 10 -Controlling User Access with Privileges 11 -Auditing Database Use and Controlling Resources and Passwords 12 -Understanding Oracle8 Backup Options 13 -Selecting and Implementing a Backup Strategy 14 -Performing Database Recovery 15 -Using Recovery Manager for Backup and Recovery 16 -Using Optimizers and the Analytic and Diagnostic Tools 17 -Using Constraints to Improve Your Application Performance 18 -Using Indexes, Clusters, Caching, and Sorting Effectively 19 -Improving Throughput with SQL, PL/SQL, and Precompilers Raman Batra 20 -Tuning Your Memory Structures and File Access 21 -Identifying and Reducing Contention 22 -Tuning for Different Types of Applications 23 -Diagnosing and Correcting Problems 24 -Configuring and Using Net8 Features

Featured Book Oracle8 Server Unleashed This book covers hot topics such as deploying on the Web; data warehousing; business objects; and integrating multimedia, HTML, Java apps, and unstructured text in databases.

http://www.informit.com/product/0789716534/ (1 of 2) [26.05.2000 16:57:39]

informit.com -- Your Brain is Hungry. InformIT - Using Oracle8 - Database - Oracle

25 -Using SQL*Loader and Export/Import 26 -Using Other Oracle8 Features and Functionality 27 -Understanding Oracle8 Options 27 -Glossary

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/product/0789716534/ (2 of 2) [26.05.2000 16:57:39]

informit.com -- Your Brain is Hungry. InformIT

Find IT Solutions Here

Exact Phrase All Words Search Tips

Book Store Upgrading and Repairing Networks, Second Edition TCP/IP Unleashed OpenCable Architecture KDE Application Development

InformIT delivers solutions to IT professionals & allows you to: Access industry news and the latest downloads. Sample content from our Free Library. Acquire new skills and search for a job. Become a member for FREE. New at InformIT Networking Explore the ins and outs of interdomain routing network designs.

Linux The second part of the Linux firewall series shows you how to configure a secure Linux firewall.

Visual Basic Bob Reselman shows you how to create a Web-aware, custom control that you can run either in a Web browser or in a Visual Basic form. Dan Appleman shows you how to utilize simple techniques to support large sites without the need to build complex and costly systems. More...

Upgrading and Repairing PCs, Linux Edition

Software Store Complete Linux O/S 7.0 Deluxe Linux O/S 7.0 Secure Server Edition Linux O/S 7.0 Maximum Security Linux Linux Utilities VB6 Programming Starter Kit The JBuilder Starter Kit

Configuration scenarios with a focus on TCP/IP, IPX, and AppleTalk network protocols.

For the final step in the firewall series, you'll install the software and finalize the configuration of your firewall. Download, install, and experiment with pre-beta versions of KDE2. More...

More...

Tell Mel... What do you listen to at work?

Take a dip in our knowledge pool. ProgramIT Java, Visual Basic, Visual C++ DatabaseIT Visual Basic, Visual C++, Oracle, SQL Server WebIT Java, Front Page, HTML, Web Design NetworkIT Linux, Windows 2000, Windows NT

MP3s The Ethics of Free Software By John Goerzen (Networking: Linux) A philosophical argument formulated around open source software development as it relates to proprietary software and closed source software development. CDs Streaming Audio Radio Coworker's conversation Other
Submit Vote

View Results

http://www.informit.com/ (1 of 3) [26.05.2000 16:57:56]

informit.com -- Your Brain is Hungry. InformIT

ConfigureIT Upgrading and Repairing, Linux, Windows 2000, Windows NT

More... Bob Reselman Bob Reselman is a nationally known software developer, writer, and teacher, as well as a Founding Principal with Cognitive Arts and Technologies.

Overview of the Microsoft Certification Process Courtesy of New Riders (Windows 2000/NT) Microsoft certifications have been in the news lately, mainly thanks to the new Windows 2000 exams. Here's where you can cut through the hype and learn the skinny on all the Microsoft certs.

Linux System Administration Linux System Administration guides the reader in the many intricacies of maintaining a secure, stable system.

What your peers read. Most Popular Book Red Hat Linux Unleashed (Networking : Linux) Most Popular Chapter Open Systems, Standards (Networking : TCP/IP)

More...

Essential Resources for XML Professionals By Benot Marchal (Programming) Benot Marchal provides a list of the books he keeps within easy reach when looking for answers on XML, as well as the web sites he regularly visit for XML updates Essential Resources for Visual Basic Professionals By Dan Fox (Programming: Visual Basic) Dan Fox provides a guide to show which Visual Basic resources he finds most useful, including books you should add to your library and VB web sites worth bookmarking..

Managing User Accounts in MySQL By Paul DuBois (Networking: Linux) The purpose of this tutorial is to explain two statements introduced in MySQL 3.22.11 that make user administration much easier: GRANT and REVOKE.

The best info on every topic.

More...

http://www.informit.com/ (2 of 3) [26.05.2000 16:57:56]

informit.com -- Your Brain is Hungry. InformIT

Visual InterDev Security (Web Technologies) Michael Marsh shows you how to protect data that is part of your site, but that is not for public viewing through controlled access. The Old Box of Variable Tricks (Programming) Don Snaith takes the mystery out of the most common C++ data types and variables.
Editors' Choice About...

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/ (3 of 3) [26.05.2000 16:57:56]

Mamma.com: The Mother of All Search Engines

eZuz.com Shopping Autos Fitness & Nutrition Find Music Picture Personals Free USA Greetings Technology Evaluation URL Submit Service Free Games @ Pogo.com Try the advanced Power Search!

Search:

The Web
Find It!

Copyright 1996-2000, Mamma.com All Rights Reserved Privacy Policy

http://www.mamma.com/ [26.05.2000 16:58:06]

informit.com -- Your Brain is Hungry. InformIT - ConfigureIT

You are here : Home : ConfigureIT Featured Author

Select a Solution Center
Exact Phrase All Words Search Tips

Bill Ball Bill Ball is a technical writer, editor, and magazine journalist, who has been working with computers for the past 20 years.

A centralized place for PC repair specialists, technicians and engineers to learn new hardware configurations, search for quick answers and solve tricky problems.

Select Topic

Featured Book Our editors have drawn upon their industry knowledge and expertise to hand-select chapters, therefore enabling you to solve problems and save time. Pre-Installation Planning Obtain the background information you'll need to begin the Windows 2000 installation process, including coverage of security and compatibility issues. From: Microsoft Windows 2000 Professional Installation and Configuration Handbook ; Jim Boyce ; Que Jobs, Processes and Threads Look closely at the internal aspects of jobs, processes and threads -- the workhorses inside the Windows 2000 OS. From: Microsoft Windows 2000 Security Handbook ; Jeff Schmidt; Dave Bixler; Theresa Hadden; Travis Davis; Dave Bixler; Alexander Kachur; Que Understanding Bits, Nybbles, and Bytes Peter Norton Explains the ways information is represented in your PC From: Peter Norton's Inside the PC, Eighth Edition ; Peter Norton; John Goodman; Sams more... Linux Mandrake O/S 7.0 Secure Server Edition The award-winning Linux-Mandrake distribution teamed with a secure Web server.

Select Topic

What's popular in our IT Reference Library? We let you tell us. Here are the most frequently viewed books and chapters for the past week. Books 1 Red Hat Linux Unleashed Timothy Parker, Sams, Linux 2 Platinum Edition Using Windows NT 4 Jerry Honeycutt, Que, Windows NT 3 Upgrading & Repairing PCs, Eighth Edition Scott Mueller, Que, Upgrading and Repairing 4 Linux SA Survival Guide , Linux Chapters 1 Introduction to Linux From:Red Hat Linux Unleashed Timothy Parker, Sams, Linux 2 Defining Your PC From:Using PCs - Pub Que, Que, General 3 Introduction From:Upgrading & Repairing PCs, Eighth Edition - Scott Mueller, Que, Upgrading and Repairing 4 From:

http://www.informit.com/configureit/ (1 of 2) [26.05.2000 16:58:14]

informit.com -- Your Brain is Hungry. InformIT - ConfigureIT

5 Using PCs Pub Que, Que, General more....

5 Motherboards From:Upgrading & Repairing PCs, Eighth Edition - Scott Mueller, Que, Upgrading and Repairing more....

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/configureit/ (2 of 2) [26.05.2000 16:58:14]

informit.com -- Your Brain is Hungry. InformIT - NetworkIT

You are here : Home : NetworkIT Featured Author

Select a Solution Center
Exact Phrase All Words Search Tips

Benot Marchal Benot Marchal runs Pineapplesoft, a consulting company that specializes in Internet applications -- particularly e-commerce, XML, and Java. In 1997, Ben co-founded the XML/EDI Group, a think tank that promotes the use of XML in e-commerce applications.

A centralized place for Network administrators and architects to study new methods, search for easy solutions and answer difficult networking questions.

Select Topic

Our editors have drawn upon their industry knowledge and expertise to hand-select chapters, therefore enabling you to solve problems and save time. Why OpenCable? Many people are asking themselves this question these days. If you're one of them, find out more about the goals of OpenCable and solutions this initiative offers. From: OpenCable Architecture ; Michael Adams ; Cisco Press (53) Mitnick Attack Get an introduction to intrusion detection by looking at one of the most famous intrusion cases that ever occurred, Kevin Mitnick's attack on Tsutomu Shimomura's system. Includes TCP/IP review and a discussion of SYN flooding. From: Network Intrusion Detection: An Analyst's Handbook ; Stephen Northcutt ; New Riders BGP Cores and Network Scalability If you're already familiar with the basic operation of BGP, read on to find out how BGP can be used to scale your network even further. From: Advanced IP Network Design (CCIE Professional Development series) ; Russ White; Alvaro Retana; Don Slice ; Cisco Press (53) more...

Featured Book Sams Teach Yourself Shell Programming in 24 Hours Take control of your systems by harnessing the power of the shell.

Select Topic

Our authors offer you a sneak peak into their books as they write them, therefore giving you the most up-to-date information available anywhere. In order to bring you AlphaBooks as quickly as possible, keep in mind they have not yet been reviewed for accuracy, that they are subject to change without notice, are presented "as-is," and are not recommended for use in critical applications where data loss would be detrimental. Computer Management-System Tools Focusing on system tools in the Computer Management snap-in, author Jim Boyce covers not only the current tool but also the Windows NT utility it replaced. From: Microsoft Windows 2000 Professional Installation and Configuration

http://www.informit.com/networkit/ (1 of 2) [26.05.2000 16:58:25]

informit.com -- Your Brain is Hungry. InformIT - NetworkIT

Handbook ; Jim Boyce ; Que NTFS 5.0 Examine in detail NTFS 5.0, the enhanced version of NTFS that ships with Windows 2000. From: Microsoft Windows 2000 Security Handbook ; Jeff Schmidt; Dave Bixler; Theresa Hadden; Travis Davis; Dave Bixler; Alexander Kachur; Que The Windows 2000 Security Model Expert author Jeff Schmidt introduces you to the schema that Windows 2000 uses to protect objects, as well as the mechanisms that enforce those protections. From: Microsoft Windows 2000 Security Handbook ; Jeff Schmidt; Dave Bixler; Theresa Hadden; Travis Davis; Dave Bixler; Alexander Kachur; Que more...

Select Topic

What's popular in our IT Reference Library? We let you tell us. Here are the most frequently viewed books and chapters for the past week. Books 1 Red Hat Linux Unleashed Timothy Parker, Sams, Linux 2 Maximum Internet Security: A Hackers Guide Publishing Sams, Sams, Intrusion Detection 3 Platinum Edition Using Windows NT 4 Jerry Honeycutt, Que, Windows NT 4 Teach Yourself TCP/IP in 14 Days, Second Edition Timothy Parker, Sams, TCP/IP 5 Linux SA Survival Guide , Linux more.... Chapters 1 Open Systems, Standards, From:Teach Yourself TCP/IP in 14 Days, Second Edition - Timothy Parker, Sams, TCP/IP 2 Introduction to Linux From:Red Hat Linux Unleashed Timothy Parker, Sams, Linux 3 TCP/IP and the Internet From:Teach Yourself TCP/IP in 14 Days, Second Edition - Timothy Parker, Sams, TCP/IP 4 From: 5 Preparing a Windows NT 4 Network for Windows 2000 From:Windows 2000 Server: Planning and Migration - Sean Deuby, MTP, Windows 2000 more....

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/networkit/ (2 of 2) [26.05.2000 16:58:25]

informit.com -- Your Brain is Hungry. InformIT - WebIT

You are here : Home : WebIT Featured Author

Select a Solution Center
Exact Phrase All Words Search Tips

Jakob Nielsen, Ph.D. Jakob Nielsen is one of the most well-respected speakers in the high-tech world because he delivers a valuable message with eloquence. His message is simple: Put usability first. Practice simplicity.

A centralized place for Webmasters, designers, and developers to learn new tricks, search for much needed answers and solve the latest online problems. Expert Recommends: Essential Resources for XML Professionals by Benot Marchal Benot Marchal provides a list of the books he keeps within easy reach when looking for answers on XML, as well as the web sites he regularly visit for XML updates Read more...

Featured Book Designing Web Usability Usability rules the web. Look at the web as you've never seen it before: through the eyes of the average user.

Select Topic

Our editors have drawn upon their industry knowledge and expertise to hand-select chapters, therefore enabling you to solve problems and save time. Manipulating Data and the Hierarchy of JavaScript Understand the hierarchy of objects and in order to conquer JavaScript successfully. From: JavaScript Goodies ; Joe Burns ; Que Using Style Sheets with Individual Pages Style sheets are a powerful tool for the design, layout, and modification of a web page. In this chapter, authors Neil Randall and Dennis Jones examine the underlying syntax of style sheet HTML and explore FrontPage 2000 tools that are used for the creation of style sheets. From: Special Edition Using Microsoft FrontPage 2000 ; Neil Randall; Dennis Jones ; Que Getting Help Ben Forta explains Tag Editor dialogs and how to use the online help system From: Sams Teach Yourself HomeSite 4 in 24 Hours ; Ben Forta ; Sams more...

Select Topic

http://www.informit.com/webit/ (1 of 2) [26.05.2000 16:58:34]

informit.com -- Your Brain is Hungry. InformIT - WebIT

What's popular in our IT Reference Library? We let you tell us. Here are the most frequently viewed books and chapters for the past week. Books 1 SE USING HTML 4, 4TH EDITION Jerry Honeycutt, Que, HTML 2 Sams Teach Yourself Java in 21 Days, Professional Reference Ed Laura Lemay; Michael Morrison, Sams, Java 3 Sams Teach Yourself JavaScript 1.1 in a Week, Second Edition Arman Danesh, Sams, JavaScript 4 Sams Teach Yourself Perl 5 in 21 Days, Second Edition David Till, Sams, Perl 5 How to Use HTML 3.2 , HTML more.... Chapters 1 An Introduction to Java Programming From:Sams Teach Yourself Java in 21 Days, Professional Reference Ed - Laura Lemay; Michael Morrison, Sams, Java 2 Where Does JavaScript Fit In? From:Sams Teach Yourself JavaScript 1.1 in a Week, Second Edition - Arman Danesh, Sams, JavaScript 3 Your First Script From:Sams Teach Yourself JavaScript 1.1 in a Week, Second Edition - Arman Danesh, Sams, JavaScript 4 Working with Data and Information From:Sams Teach Yourself JavaScript 1.1 in a Week, Second Edition - Arman Danesh, Sams, JavaScript 5 Object-Oriented Programming and Java From:Sams Teach Yourself Java in 21 Days, Professional Reference Ed - Laura Lemay; Michael Morrison, Sams, Java more....

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/webit/ (2 of 2) [26.05.2000 16:58:34]

informit.com -- Your Brain is Hungry. InformIT - DatabaseIT

You are here : Home : DatabaseIT Featured Author

Select a Solution Center
Exact Phrase All Words Search Tips

Megh Thakkar Megh Thakkar is the Director of Database Technologies at Quest Software in Australia. Previously, he worked as a technical specialist at Oracle Corporation. Megh is the author of E-Commerce Applications Using Java and Oracle 8I From Scratch (Que, 2000).

A centralized place for database developers, designers, and administrators to learn new data models, search for answers to difficult questions, and solve database problems. Tutorial: Using AUTO_INCREMENT to Create Sequences by Paul DuBois The following tutorial discusses the function of the AUTO_INCREMENT variable and the way it works with sequences. Read more...

Featured Book Oracle8 Server Unleashed This book covers hot topics such as deploying on the Web; data warehousing; business objects; and integrating multimedia, HTML, Java apps, and unstructured text in databases.

Managing User Accounts in My SQL The purpose of this tutorial is to explain two statements introduced in MySQL 3.22.11 that make user administration much easier: GRANT and REVOKE By Paul DuBois

Select Topic

Our editors have drawn upon their industry knowledge and expertise to hand-select chapters, therefore enabling you to solve problems and save time. SQL Server Installation SQL Server 7 is almost as easy as running a setup. Before you turn the key, take a look at what's under the hood and what it takes to power it. From: Microsoft SQL Server 7.0 Unleashed ; Simon Gallagher; Sharon Bjeletich; Vipul Minocha; Greg, et al Mable; David Solomon; Matt Shepker; Tibor Karaszi; Kim Tripp; Robert Pfeiff; Joe Giordano; Jeff Knutson; Edgard Luque; David Winters; Patrick Galluci; Fulcrum Software; Irfan Chaudhry; Kevin Viers; Sams New User Features in Access 2000 Get to know the new features of Access 2000 From: Using Microsoft Access 2000 ; Susan Harkins; Tom Gerhart; Ken Hansen ; Que Exploring The Data Foundations--The Table Join Paul Cassel for some real, hands-on use of Access, including creating tables and editing table properties. From: Sams Teach Yourself Microsoft Access 2000 in 21 Days ; Paul Cassel; PALMER CONSULTING; Sams

http://www.informit.com/databaseit/ (1 of 2) [26.05.2000 16:58:44]

informit.com -- Your Brain is Hungry. InformIT - DatabaseIT

more...

Select Topic

What's popular in our IT Reference Library? We let you tell us. Here are the most frequently viewed books and chapters for the past week. Books Chapters 1 Sams Teach Yourself SQL in 21 1 Introducing Relational Databases Days, Second Edition and Oracle8 Bryan Morgan, Sams, SQL From:Using Oracle8 - David Austin, Que, Oracle 2 Using Oracle8 David Austin, Que, Oracle 2 Introduction to SQL Server 6.5 and Relational Databases 3 Waite Group's Visual Basic 6 From:Sams Teach Yourself Database How-To Microsoft SQL Server 6.5 in 21 Days Eric Winemiller, Sams, Visual Basic - Rick Sawtell; Lance Mortensen; Richard Waymire, Sams, SQL Server 4 Oracle Unleashed 3 Introduction to the Query: The Ais Et al, Sams, Oracle SELECT Statement 5 Sams Teach Yourself Oracle8 From:Sams Teach Yourself SQL in in 21 Days 21 Days, Second Edition - Bryan Ed Whalen, Sams, Oracle Morgan, Sams, SQL more.... 4 Introduction to SQL From:Sams Teach Yourself SQL in 21 Days, Second Edition - Bryan Morgan, Sams, SQL 5 Accessing a Database with Bound Controls From:Waite Group's Visual Basic 6 Database How-To - Eric Winemiller, Sams, Visual Basic more....

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/databaseit/ (2 of 2) [26.05.2000 16:58:44]

informit.com -- Your Brain is Hungry. InformIT - ProgramIT

You are here : Home : ProgramIT Featured Author

Select a Solution Center
Exact Phrase All Words Search Tips

Rhys Lewis Rhys Lewis is a senior software architect at Candle Corporation, where he works on products and services for eBusiness Application Integration(eAI). Rhys has more than 20 years of experience in the IT industry, where he has primarily worked in software design, development, and consulting.

A centralized place for developers of all experience levels to learn new coding tricks, search for answers to difficult problems, and solve key programming issues. Essential Resources for Software Development Professionals by Bob Reselman Bob Reselman provides a list of the resources he uses for all aspects of software development, including ASP, VBScript, Visual Basic, JavaScript, Windows DNA, object-oriented programming, and the art and practice of software development in general. Read more...

Featured Book Advanced Messaging Applications with MSMQ and MQSeries Advanced MSMQ programming is written for advanced Windows 95/98/NT developers and C programmers who need to know the architecture and programming concepts associated with the Microsoft Message Queue Server.

New Namespaces Namespaces are a capable and yet intricate feature that was added to C++ several years ago. Today every non trivial C++ application uses this feature either directly or indirectly (for example, by referring to functions and classes defined in the Standard Library). This article explains the rationale behind namespaces, their usage, and interaction with other C++ constructs. By Danny Kalev New Data Management Alternatives for IIS (webclass) and ASP Application Servers MTS and databases are commonly used to build Internet applications, but they may be overkill for someas you'll see in the Internet discussion group application presented here. By Daniel Appleman

New Users of Message Queuing Technologies Rhys Lewis guides you to the best resources for message queueing, including IBMs MQSeries and Microsofts MSMQ. By Rys Lewis Essential Resources for XML Professionals Benot Marchal provides a list of the books he keeps within easy reach when looking for answers on XML, as well as the web sites he regularly visits for XML updates. By Benot Marchal

http://www.informit.com/programit/ (1 of 3) [26.05.2000 16:58:55]

informit.com -- Your Brain is Hungry. InformIT - ProgramIT

Select Topic

Our editors have drawn upon their industry knowledge and expertise to hand-select chapters, therefore enabling you to solve problems and save time. The GUIDS Methodology for Object-Oriented Architecture Get a description of the GUIDS Methodology for object-oriented design (OOD). From: Doing Objects in Visual Basic 6 ; Deborah Kurata ; Sams Designing Message Queuing Applications Message queuing products provide powerful mechanisms for connecting applications. However, this very flexibility also means added design as well as implementation difficulty. This chapter guides you through the maze. From: Advanced Messaging Applications with MSMQ and MQSeries ; Rhys Lewis ; Que Installing the JDK and Getting Started Joseph Weber takes you through each step of the JDK installation process and gets you started with the Java 2 platform. From: Special Edition Using Java 2 Platform ; Joseph Weber ; Que more...

Select Topic

Our authors offer you a sneak peak into their books as they write them, therefore giving you the most up-to-date information available anywhere. In order to bring you AlphaBooks as quickly as possible, keep in mind they have not yet been reviewed for accuracy, that they are subject to change without notice, are presented "as-is," and are not recommended for use in critical applications where data loss would be detrimental. Starting the Project Learn how to understand requirements and formulating a design with Visual Basic 6. From: Visual Basic 6 From Scratch ; Robert P. Donald; Gabriel Oancea ; Que The Active Directory Operational Attributes Each Active Directory domain controller maintains a set of information that describes the state and configuration of the domain controller itself. Explore this information with author Gil Kirkpatrick. From: Active Directory Programming ; Gil Kirkpatrick ; Sams The Components of Active Directory Get the basics from Gil Kirkpatrick by reading about the components of Active Directory. A must read! From: Active Directory Programming ; Gil Kirkpatrick ; Sams more...

Select Topic

http://www.informit.com/programit/ (2 of 3) [26.05.2000 16:58:55]

informit.com -- Your Brain is Hungry. InformIT - ProgramIT

What's popular in our IT Reference Library? We let you tell us. Here are the most frequently viewed books and chapters for the past week. Books 1 Using Visual Basic 6 Bob Reselman, Que, Visual Basic 2 Sams Teach Yourself C++ in 21 Days, Second Edition Jesse Liberty, Sams, C++ 3 Special Edition Using Visual C++ 6 Kate Gregory, Que, Visual C++ 4 Java Unleashed, Second Edition Michael Morrison, Java 5 Sams Teach Yourself C in 21 Days, Fourth Edition Peter Aitken, Sams, C more.... Chapters 1 Building Your First Application From:Using Visual Basic 6 - Bob Reselman, Que, Visual Basic 2 Getting Started From:Sams Teach Yourself C++ in 21 Days, Second Edition - Jesse Liberty, Sams, C++ 3 The Parts of a C++ Program From:Sams Teach Yourself C++ in 21 Days, Second Edition - Jesse Liberty, Sams, C++ 4 Building Your First Windows Application From:Special Edition Using Visual C++ 6 - Kate Gregory, Que, Visual C++ 5 Pointers From:Sams Teach Yourself C++ in 21 Days, Second Edition - Jesse Liberty, Sams, C++ more....

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/programit/ (3 of 3) [26.05.2000 16:58:55]

InformIT : Search Tips

You are here : Home : Search Tips Featured Author

Search Tips
Exact Phrase All Words Search Tips

Benot Marchal Benot Marchal runs Pineapplesoft, a consulting company that specializes in Internet applications -- particularly e-commerce, XML, and Java. In 1997, Ben co-founded the XML/EDI Group, a think tank that promotes the use of XML in e-commerce applications.

InformIT's search helps you find content, reference, and related information on InformIT. It is not a resource to search the entire Internet. Here's how it works:

1. Type your keyword(s) in the search box. You must use specific keywords and not complete sentences or sentence fragments. Entering: program Will find: program, programmer, programming, and similar words BUT Entering: How can I become a programmer? Will find: Various, unrelated topics 2. Use the radio buttons in the search box to select Exact Phrase or All Words. 3. Press the Search button to start your search. 4. If a Basic Search does not meet your needs, try Using Boolean Expressions. Exact Phrase Search The Exact Phrase search tells the search engine to find your exact keywords words in sequence. Entering: visual basic programming Would Find: Visual Basic Programming in 12 Easy Lessons and similar terms All Words Search The All Words search looks for your exact keywords but not necessarily in sequence. Entering: visual basic programming Would Find: programming with visual basic, programming basics of visual c++, spreadsheet basics, shell programming and similar terms

Featured Book Linux System Administration Linux System Administration guides the reader in the many intricacies of maintaining a secure, stable system.

Basic Search Almost everything you need to search for can be found quickly and with better results using the standard search box, where the InformIT search service sorts results by placing the most relevant content first. Here are a few guidelines to get you started:
q

Check Spelling Make sure your search terms are spelled correctly. Use a comma to separate terms Documents found must contain the first term and variations of the second term. HTML, editor will search for html and variations of editor.

http://www.informit.com/search/tips/ (1 of 2) [26.05.2000 16:59:04]

InformIT : Search Tips

q

Use more than one word In general, the more specific you are, the better your results will be. A search for HTML 4 will produce better results than a search for languages. Be specific Use specific words as opposed to general ones. For example, a search for C++ will return more targeted results than a search for"programming.

Using Boolean Search Expressions Boolean search is for very specific searches and not for general searching. However, if you have to do some complex Boolean searches our search mechanism can meet your needs. Here is a list of the Boolean searches InformIT is capable of as well as the search type that supports them:
q

"Quotation Marks" Documents found must contain the exact phrase in "quotation marks". "Visual Basic" finds documents where the exact search of visual basic appears. This option works for both the Exact Phrase and All Words searches. AND Documents found must contain all words joined by the AND operator. Visual AND Basic finds documents with both the word visual and the word basic. This option works only in the Exact Phrase search. OR Documents found must contain at least one of the words joined by OR. Visual OR Basic finds documents containing either visual or basic. The found documents could contain both, but do not have to. This option works only in the Exact Phrase search. NOT Documents found cannot contain the word that follows the term NOT. Visual NOT Basic finds documents with visual but not containing basic. This option works only in the Exact Phrase search. ? Documents found will contain the letters that follow or precede the ?. C? finds Visual C+, Visual C++, C, and similar words. This option works only in the Exact Phrase search. * Documents found will find documents containing characters of any length. Progra* finds documents containing program, programmer, programming, and similar words. This option works only in the Exact Phrase search.

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 1999-2000 Macmillan USA. All rights reserved. InformIT is a trademark of Macmillan USA.

http://www.informit.com/search/tips/ (2 of 2) [26.05.2000 16:59:04]

InformIT : Dice.com Jobs

You are here : Home : Dice.com Jobs

Enter job search keywords:

Search Find all of the words Find Any of the words Use the Boolean expression Type of Employment:
No Restrictions Contract - W2 Contract - Independent Contract - Corp-to-Corp Contract to Hire - W2 Contract to Hire - Independent Contract to Hire - Corp-to-Corp Full time
Hold CTRL to select multiple options

Clear

Telephone Area Codes:

Job Location: All US states Travel Telecommute International Alabama Alaska Arizona Arkansas California Canada

Display only jobs that were posted: In the last 30 days

Search Tip: Good keywords include Computer Languages & Protocols, Software Packages, Hardware Platforms, Functional Job Titles, Company Names, State Abbreviations, City names, Telephone Area Codes, State Zip codes. For example:

Sacramento CA UNIX 916

http://www.informit.com/jobs/ [26.05.2000 16:59:12]

Your Brain is Hungry. InformIT.

You are here : Home : Careers Featured Author

Careers at InformIT
Exact Phrase All Words Search Tips

Benot Marchal Benot Marchal runs Pineapplesoft, a consulting company that specializes in Internet applications -- particularly e-commerce, XML, and Java. In 1997, Ben co-founded the XML/EDI Group, a think tank that promotes the use of XML in e-commerce applications.

Wanted: Brain Surgeons If you operate in the Information Technology world and have the pedigree to delve deep into the human mind, read on. We are M.D.'s of IT and physicians of fast answers with the right prescription for information needs. We heal through online content, reference, and training. We cure the common question, remove doubt from clouded minds, and make right-brains right again. We train experts in the areas of programming, databases, Web publishing, networking, and operating systems, and draw from world's largest compilation of technical computer reference content.

Your Bank Account is Hungry... Join us!

We have entry-level, management, and senior level positions available in Human Resources, Web Development, Business Development, and Marketing. If you're interested and you're smarter than the average Ph.D., join our operating team. You'll receive medical, dental, vision, 401K, plus healthy advancement opportunities. Think of the challenges, the fun, and all of the suffering that you'll relieve. E-mail your resume to [email protected] or snail mail it to InformIT, P.O. Box 80015, Indianapolis, IN, 46280-0015. IT begins with diversity - an equal opportunity employer. Project Manager Customer Service Manager Public Relations Manager Web Development Professional Director of Business Development XML "Geek"

Featured Book Linux System Administration Linux System Administration guides the reader in the many intricacies of maintaining a secure, stable system.

Project Manager
Seeking detail-oriented project leader who is adept at multi-tasking, decision-making, and stimulating creativity within a team environment. This high-profile position requires excellent communication skills and market insight/savvy. Must possess a clear understanding of computer technology, the end user/customer, the IT industry, and Internet/e-commerce models. E-mail your resume to [email protected]

Customer Service Manager

This team-leadership position requires collaboration with online developers and managers as well as communication with non-technical employees in other inter-company business groups. Qualifications include customer service experience, working knowledge of Internet technologies, excellent oral and written communication skills, and experience dealing with internal and external customers. Corporate help desk or system administration experience a plus. E-mail your resume to [email protected]

http://www.informit.com/careers/ (1 of 2) [26.05.2000 16:59:22]

Your Brain is Hungry. InformIT.

Web Development Professional 1

Open systems development on UNIX or Linux with experience in two or more of the following languages or technologies: Oracle, SQL, HTML, JavaScript, Perl or Cold Fusion. E-mail your resume to [email protected]

Web Development Professional 2

Excellent problem-solving and algorithm skills are needed in a fast-paced environment. Experience and demonstrated skills in one or more of the following languages or technologies: SQL, SQL Server, Visual Basic 6.0, Cold Fusion, HTML or JavaScript. E-mail your resume to [email protected]

Public Relations Manager

Build and manage a strategic, aggressive, and focused PR program for a newly-formed, high-tech IT company. This high-profile position requires a motivated self-starter and a strong communicator with proven experience overseeing day-to-day PR activities and managing an outside PR firm. E-mail your resume to [email protected]

Director of Business Development

Serve as the integral link between third party sites and future online partners for a Web-driven company. You'll be responsible for establishing, negotiating and overseeing strategic partnerships with high-profile, online content sites, e-commerce merchants, and offline computer information providers. E-mail your resume to [email protected]

XML "Geek"
Needed: An experienced web developer who has hands-on experience with the XML revolution. You've built your own web sites and live and breathe the 'Net. q XML, DTDs and XSL q Document formats and processing q Database integration with web sites q Unix admin and support q Familiarity with DocBook and DTD a plus q Apache module development desirable q Experience with Perl, Java and Cocoon a plus E-mail your resume to [email protected]

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/careers/ (2 of 2) [26.05.2000 16:59:22]

InformIT : Press Area

You are here : Home : Press Area Featured Author

Exact Phrase All Words Search Tips

Welcome to InformIT's Press Area

InformIT is a complete online community that delivers information, technology, reference, training, news, and opinion to IT professionals, students, and corporate users. Rely on this area for press releases, event coverage, background materials, screen shots, and logo specifications. For further information, please contact [email protected] Press Releases Find out the latest happenings at InformIT. Facts About InformIT All the information you'll need about InformIT's past, present, and future. Logo Specifications/Screen Shots You can obtain logo specifications and screen shots of the InformIT site by contacting [email protected].

Benot Marchal Benot Marchal runs Pineapplesoft, a consulting company that specializes in Internet applications -- particularly e-commerce, XML, and Java. In 1997, Ben co-founded the XML/EDI Group, a think tank that promotes the use of XML in e-commerce applications.

Top Ten Reasons There are countless reasons to visit InformITwe'll start you off with the top ten.

Featured Book Linux System Administration Linux System Administration guides the reader in the many intricacies of maintaining a secure, stable system.

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/press/ [26.05.2000 16:59:28]

InformIT : About Us

You are here : Home : About InformIT Featured Author

About InformIT
Exact Phrase All Words Search Tips

Benot Marchal Benot Marchal runs Pineapplesoft, a consulting company that specializes in Internet applications -- particularly e-commerce, XML, and Java. In 1997, Ben co-founded the XML/EDI Group, a think tank that promotes the use of XML in e-commerce applications.

You've arrived at an integrated environment that contains the largest compilation of content and services for IT professionals and students. At InformIT, you will find a place to share information through interaction with your peers and experts as well as a place to discover solutions to all your IT problems. We are hopeful that InformIT will become your top destination for IT solutions! Facts About InformIT Take a Tour All the information you'll need about Helpful tips for navigating InformIT. An interactive InformIT's past, present, and future. overview is available (requires Feedback/Comments Shockwave 4). Let us know what's on your mind...we're listening. Register If you aren't yet registered, Careers at InformIT sign up now to gain access to InformIT is in need of a few good all of InformIT's free content! minds. Advertising If you are interested in advertising on InformIT, we'd love to speak with you!

Featured Book Linux System Administration Linux System Administration guides the reader in the many intricacies of maintaining a secure, stable system.

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 1999-2000 Macmillan USA. All rights reserved. InformIT is a trademark of Macmillan USA.

http://www.informit.com/about/ [26.05.2000 16:59:34]

Your Brain is Hungry. InformIT.

You are here : Home : ZD Network News

ZD Network News for 5/26/2000

Exact Phrase All Words Search Tips

---

DIGNAN
There's a lot of good news about Linux -- too bad Wall Street doesn't want to hear it.

WEB SITE KILLS MURDER TRIAL

4:40 AM PT Aussie site threatened with 'serious sanctions' after it causes judge to abort murder retrial. They're 'bullying us,' CrimeNet says.

ALPHA CHIP UPDATE -- IT AIN'T DEAD YET

4:35 AM PT In fact, the 64-bit chip that fuels high-end servers, workstations and e-commerce is undergoing a renaissance -- with 1GHz just around the bend.

MICROSOFT-DOJ TRIAL: WHAT'S COMING NEXT?

The court battle has dragged on for two years now, but the case is far from over. Find out what the next steps will be.

DOTCOM MAKING WAVES WITH TRAVEL OFFER

Biztravel.com offers refunds for late or canceled flights. Too good to last, say analysts, but company stands by its claim.

INTERNET MUSIC -- IN THE LINE OF FIRE

Congressional hearings, new lawsuits from Snoop Dogg's record label and a study claiming college piracy is hurting music sales put MP3 downloads in the spotlight.

JUDGE PROTECTS EBAY FROM CONTENT RAIDS

Bidder's Edge ordered to stop scouring the auction giant's site using robots, crawlers and the like, in search of auction listings.

COFFEE: MS CASE DISTRACTS US FROM THE FUTURE FOLEY

Microsoft has attempted to equate the ill-defined term 'integration' with 'innovation.'

TRADE GROUPS STEAL ANTITRUST SHOW

MAY 24 Is the future of Microsoft contained in a little-known legal filing from two tech industry groups? It's certainly caught the judge's eye.

COOP: JUDGE PREFERS THREESOME

MAY 24 Why The Hanging Judge may be tripping over his own noose.

DVORAK: I'LL ADMIT I'VE MADE A FEW BAD CALLS ALONG THE WAY.
I'm man enough to admit I've made a few bad calls along the way.

MS REMEDY DAY NEARS

MAY 23 Wednesday is R-day -- remedy hearing day -- for Microsoft and the government in the antitrust trial of the new millennium, and almost anything can happen.

http://www.informit.com/news/ (1 of 2) [26.05.2000 16:59:49]

Your Brain is Hungry. InformIT.

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising 2000 InformIT. All rights reserved.

http://www.informit.com/news/ (2 of 2) [26.05.2000 16:59:49]

InformIT : Download It

You are here : Home : Download It

About the Partners

Go to....

SEARCH:

Upload

Contact

Home

Start Here:
Software Head of the Herd Newsletter Top 100 Search Index What's New

Head of the Herd

Mailbag Assistant

BeOS Java Windows 3x Windows NT Windows 95/98 Windows 2000

One benefit of email, compared to handwritten letters, is that you have a history of what was said. You can use that history in many ways, if you can find what you're looking for. Read our full review!

TUCOWS NETWORK

Affiliates Domain Direct Games HTMLStuff Kickoff Linux LWN ISP Central Macintosh Music Network News OpenSRS PDA

Hot Software

TUCOWS News

http://www.informit.com/downloads/ (1 of 3) [26.05.2000 17:00:05]

InformIT : Download It
Shop Themes TUCOWS TUKIDS UNIX Themes

First Page 2000 is a HTML/script authoring tool for professional and blooming webmasters. This powerful program comes with an eye-catching interface and everything you need to create professional websites. Read Our Review With new INFRANET 2000 you will increase your productivity because of not wasting your time on searching the web or opening dozen of WEB BROWSER windows. Search the entire internet with more than 240 search engines and do it faster using multiweb searching! Read Our Review

New License Type TUCOWS introduces a NEW License Type: Adware. By popular demand, we have decided to split our freeware section into two parts, Freeware & Adware. Both will remain free to use for an unlimited time. The difference is that Adware programs are banner supported software. Shareware Awards
Choose your favorite software! That's right, it is that time again: The Shareware Industry Awards! Head on to http://www.tucows.com/sic.html and vote now!

Software Links
q q q q q q

Help, Tips and Tricks

TUCOWS is dedicated to providing the best possible service on the Internet. Our staff has compiled a list of Helpful Information, tips, facts, and frequently asked questions that may be just what you are looking for.

Windows 95/98 Windows NT Windows 2000 BeOS Java Windows 3.x

Mailing List
Keep up-to-date on all the latest happening at TUCOWS! Join our mailing list and recieve updates on our featured programs, news, new software added and the latest updated software. To find out how, moove on over to our Signup Page!

About TUCOWS
Want to learn more about TUCOWS? Head on to About TUCOWS! This site has information on those that make this company run. Plus, a collection of Press releases that reveals the projects that we are working on right now! Also, learn about just how many people come to TUCOWS anyhow.

With over 1000 Mirror Sites worldwide, TUCOWS provides the FASTEST Internet downloads ANYWHERE! Our global network of Internet servers is also so powerful, that companies such as ICQ & Netscape, (among others) distribute their products, directly through the TUCOWS Mirror Network.

http://www.informit.com/downloads/ (2 of 3) [26.05.2000 17:00:05]

InformIT : Download It

News

Contact

Advertise

CD ROM

Search

Upload

Help

Home

Copyright 2000 TUCOWS.Com Inc. TUCOWS.Com Inc. has no liability for any content or goods on the TUCOWS site or the Internet, except as set forth in the terms and conditions and privacy statement.

http://www.informit.com/downloads/ (3 of 3) [26.05.2000 17:00:05]

informit.com -- Your Brain is Hungry. InformIT - Buy It

You are here : Home : Buy It Featured Author Benot Marchal

Exact Phrase All Words Search Tips

Categories CAD AutoCAD 14 AutoCAD 2000 Certification CCIE MCSE CCNA CCDA Other more...

Creative Database Photoshop General Illustrator Other Access SQL Server SAP Visual Macromedia more... Basic Oracle more... Desktop Applications Finance and Business Office 2000 Office SBE Excel Word Quickbooks Quicken TurboTax 2000 PowerPoint more... Investing Money more... Games Playstation Multi-platform N64 Action Strategy more... Networking Cisco IOS and Protocols Windows NT Other Linux Windows 2000 more... Programming Visual Basic Windows C++ Java Visual C++ more... Using the Internet/Web Internet Explorer Netscape Communicator General Internet AOL Hardware General Upgrading and Repairing PalmPilot iMac Dictionary more... Operating Systems Windows 98 Macintosh Windows 95 General SystemWorks more... Software Linux Web Authoring Palm Pack Add-ons Webster's -- Language more... Web Technologies HTML Java Web Business FrontPage XML more...

Benot Marchal runs Pineapplesoft, a consulting company that specializes in Internet applications -- particularly e-commerce, XML, and Java. In 1997, Ben co-founded the XML/EDI Group, a think tank that promotes the use of XML in e-commerce applications.

Featured Book Linux System Administration Linux System Administration guides the reader in the many intricacies of maintaining a secure, stable system.

Order A Book
Once you have found the book you're interested in, click on the "Order this book" link and fill out the form. Then choose one of the following methods: q Online/E-mail
q q

Phone/Fax Mail

Customer Service Information

Our customer service department is open from 7:30 a.m. to 6:30 p.m. Central time (USA), Monday through Friday. At other times, calls are answered by our voice mail. Customers in the US may call us at 1-800-716-0044. International customers may reach us at 1-317-228-4366. Our customer service department can also be reached through e-mail. Please
http://www.informit.com/store/ (1 of 2) [26.05.2000 17:00:19]

informit.com -- Your Brain is Hungry. InformIT - Buy It

send your customer service inquiries to [email protected] Order Status If you have a question regarding the status of your order, we will respond to you promptly through e-mail. Report Physical Defects If you have received a damaged or defective product, we will be glad to replace the product. Return Policy Please click on the link to review our return policy.

Shipping Information
If your order is in stock, we will ship it within 48-72 hours after you place your order. If you live in the US, you should have the book within seven days. The default domestic shipping method on our order form is Airborne Express 2nd day. If you are ordering from within the US, other shipping options are available. Only UPS International is used to ship products outside of the US. International shipping may take as long as three weeks. Shipping Charges

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/store/ (2 of 2) [26.05.2000 17:00:19]

informit.com -- Your Brain is Hungry. InformIT - Free Library

You are here : Home : Free Library InformIT Poll Have you used wireless technology to access the Internet? Yes Categories CAD AutoCAD 14 Database Access SQL Server Visual Basic Oracle ABAP more... Finance and Business Quicken Creative Photoshop Illustrator CorelDRAW Desktop Applications Excel Publisher Works Lotus Notes Word 97 more... Hardware General Upgrading and Repairing No
Submit Vote

Exact Phrase All Words Search Tips

View Results

Featured Author Benot Marchal Benot Marchal runs Pineapplesoft, a consulting company that specializes in Internet applications -- particularly e-commerce, XML, and Java. In 1997, Ben co-founded the XML/EDI Group, a think tank that promotes the use of XML in e-commerce applications.

Networking Operating Systems Windows NT Linux TCP/IP Windows 98 Windows 95 UNIX Routing and Routers more... Programming Visual Basic Windows C++ Java Visual C++ more... Web Technologies HTML Java FrontPage Visual InterDev Perl more... All Titles in the Free Library Alphabetical List Using the Internet/Web Netscape Communicator General Internet

Featured Book Linux System Administration Linux System Administration guides the reader in the many intricacies of maintaining a secure, stable system.

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/itlibrary/ [26.05.2000 17:00:26]

informit.com -- Your Brain is Hungry. InformIT - InformIT - Editors' Choice

You are here : Home :InformIT - Editors' Choice Featured Author Benot Marchal
Exact Phrase All Words Search Tips

Our editors have drawn upon their industry knowledge and expertise to hand-select chapters, therefore enabling you to solve problems and save time.

Certification

Select Topic

Implementation Build a plan of action for successfully installing, configureing, and maintaining a network From: MCSE TestPrep: Networking Essentials, Second Edition ; Jay Adamson ; New Riders PL/SQL in Use Extend the concept of modularity and top-down design by exploring the subprograms, stored subprograms, and packages in PL/SQL. From: OCP Training Guide: Oracle DBA ; Willard Baird ; New Riders Installation and Configuration Installation of operating systems and application software on client desktops is one of the most time-consuming activities performed by network administrators. Here Dennis Maione teaches how to install Windows NT Workstation 4 and ensure that all components from adapter cards to device drivers are running smoothly. From: MCSE Training Guide: Windows NT Workstation 4, Second Edition ; Dennis Maione ; New Riders more... Database

Benot Marchal runs Pineapplesoft, a consulting company that specializes in Internet applications -- particularly e-commerce, XML, and Java. In 1997, Ben co-founded the XML/EDI Group, a think tank that promotes the use of XML in e-commerce applications.

Featured Book Linux System Administration Linux System Administration guides the reader in the many intricacies of maintaining a secure, stable system.

Select Topic

Windows DNA AND COM By encapsulating legacy code in components and using DNA to guide new development, you can have the best of the client-server world by adding Internet elements (scripting, a ubiquitous platform, and reusable components). Randy Abernethy shows you how. From: COM/DCOM Unleashed ; Randy Abernethy; Randy Morin; Jesus Chahin; Sams Advanced Database Management Concepts No other single factor has a greater influence on the success of a database application than the design of the database itself. FoxPro expert Menachem Bazian gives you'll the foundation you need for successful development. From: Special Edition Using Visual FoxPro 6 ; Menachem Bazian ; Que Security and Licensing In-depth coverage of the security architecture of SQL-Server 7. From: SQL Server System Administration ; Chris Miller; Sean Baird; Michael Hotek; Denis Darveau; John Lynn; New Riders more...

http://www.informit.com/edchoice/ (1 of 3) [26.05.2000 17:00:40]

informit.com -- Your Brain is Hungry. InformIT - InformIT - Editors' Choice

Hardware

Select Topic

Networking

Select Topic

Managing Enterprise Services Author Steven B. Thomas of Service Management discusses the concept of the service process and how it differs from a regular application process that uses its own security context. From: Windows NT Heterogeneous Networking ; Steven Thomas ; MTP Object-Oriented User Interface Development This chapter places OVID within the software development life cycle and introduces the phases of the OVID methodology. Great information for the team handling this complex tasks. From: Designing for the User with OVID ; Scott Isensee; Dick Berry; John Mullaly; Dave Roberts ; MTP Basic Network Management Rick Sturm explains the basics the many challenges that network managers face. From: Working with Unicenter TNG ; Rick Sturm ; Que more... Operating Systems

Select Topic

Programming

Select Topic

Installing the JDK and Getting Started Joseph Weber takes you through each step of the JDK installation process and gets you started with the Java 2 platform. From: Special Edition Using Java 2 Platform ; Joseph Weber ; Que Allowing User Interaction--Integrating The Mouse Author Jeff Heaton takes you through the steps of allowing user interaction by integrating the mouse. From: Sams Teach Yourself Visual C++ 6 in 21 Days, Professional Reference Edition ; JEFF HEATON; Davis Chapman; Sams The Old Box of Variables Tricks Don Snaith takes the mystery out of the most common C++ data types and variables. From: Complete Idiot's Guide to C++ ; Paul Snaith ; Que more... Using the Internet/Web

Select Topic

Web Technologies

Select Topic

http://www.informit.com/edchoice/ (2 of 3) [26.05.2000 17:00:40]

informit.com -- Your Brain is Hungry. InformIT - InformIT - Editors' Choice

Visual InterDev Security Michael Marsh shows you how to protect data that is part of your site, but that is not for public viewing through controlled access. From: Visual InterDev 6 Unleashed ; Paul Thurrott ; Sams MVC Architecture Understanding the Model View Controller (MVC) architecture will enhance your ability to build robust JFC applications. From: JFC Unleashed ; Michael Foley ; Sams MAKING APPLETS LIVE ON THE WEB Rick Leinecker spend the day walking you through making your Java applets live on a Web site From: Sams Teach Yourself Visual J++ 6 in 21 Days ; Rick Leinecker ; Sams more...

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/edchoice/ (3 of 3) [26.05.2000 17:00:40]

informit.com -- Your Brain is Hungry. InformIT - AlphaBooks

You are here : Home :AlphaBooks Featured Author Benot Marchal

Exact Phrase All Words Search Tips

Our authors offer you a sneak peak into their books as they write them, therefore giving you the most up-to-date information available anywhere. In order to bring you AlphaBooks as quickly as possible, keep in mind they have not yet been reviewed for accuracy, that they are subject to change without notice, are presented "as-is," and are not recommended for use in critical applications where data loss would be detrimental.

Networking

Select Topic

Computer Management-System Tools Focusing on system tools in the Computer Management snap-in, author Jim Boyce covers not only the current tool but also the Windows NT utility it replaced. From: Microsoft Windows 2000 Professional Installation and Configuration Handbook ; Jim Boyce ; Que NTFS 5.0 Examine in detail NTFS 5.0, the enhanced version of NTFS that ships with Windows 2000. From: Microsoft Windows 2000 Security Handbook ; Jeff Schmidt; Dave Bixler; Theresa Hadden; Travis Davis; Dave Bixler; Alexander Kachur; Que The Windows 2000 Security Model Expert author Jeff Schmidt introduces you to the schema that Windows 2000 uses to protect objects, as well as the mechanisms that enforce those protections. From: Microsoft Windows 2000 Security Handbook ; Jeff Schmidt; Dave Bixler; Theresa Hadden; Travis Davis; Dave Bixler; Alexander Kachur; Que more... Programming

Benot Marchal runs Pineapplesoft, a consulting company that specializes in Internet applications -- particularly e-commerce, XML, and Java. In 1997, Ben co-founded the XML/EDI Group, a think tank that promotes the use of XML in e-commerce applications.

Featured Book Linux System Administration Linux System Administration guides the reader in the many intricacies of maintaining a secure, stable system.

Select Topic

The Active Directory Operational Attributes Each Active Directory domain controller maintains a set of information that describes the state and configuration of the domain controller itself. Explore this information with author Gil Kirkpatrick. From: Active Directory Programming ; Gil Kirkpatrick ; Sams The Components of Active Directory Get the basics from Gil Kirkpatrick by reading about the components of Active Directory. A must read! From: Active Directory Programming ; Gil Kirkpatrick ; Sams Lightweight Directory Access Protocol (LDAP) Fundamentals Gil Kirkpatrick discusses Lightweight Directory Access Protocol (LDAP) fundamentals, and more specifically, the Microsoft implementation of that API. From: Active Directory Programming ; Gil Kirkpatrick ; Sams

http://www.informit.com/alphabook/ (1 of 2) [26.05.2000 17:00:52]

informit.com -- Your Brain is Hungry. InformIT - AlphaBooks

more...

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/alphabook/ (2 of 2) [26.05.2000 17:00:52]

The Linux Infobase

What is the Linux Infobase?

Member Login Username Password

This Infobase is a searchable collection of Log In Linux content from InformIT. You'll benefit Forgot your password? from volumes of books rather than just a few random chapters. In addition, we've linked together all the Infobase chapters for quick and easy cross-referencing. With InformIT's Linux Infobase, you'll get the latest Linux content from the leading publishers easily a $300.00 value for only $14.99 a month. Now that's a deal! Subscribe now!

Features
The Linux Infobase contains premium and extensively-linked content with many helpful features: q An Advanced Search feature for searches with filtered or specific terms. Try the search before you buy.
q q

Code examples and exercises to download. Quick Links and Automatic Searches to direct you to the most frequently accessed information. Seamless interaction with MyInformIT for further customization. Content from leading technical publishers here's what you'll get.

q q

Subscription Information
InformIT's Linux Infobase costs $14.99 per month, or you can buy a yearly subscription for $99.95 that's an additional 50% savings. Don't delay! Subscribe now and enjoy your very own Linux Infobase where and how you want it!

http://infobase.informit.com/infobase/linux/informit/ (1 of 2) [26.05.2000 17:01:06]

The Linux Infobase

Contact Us | Copyright, Terms & Conditions | Privacy Policy Copyright 1999-2000 Macmillan USA. All rights reserved. InformIT is a trademark of Macmillan USA.

http://infobase.informit.com/infobase/linux/informit/ (2 of 2) [26.05.2000 17:01:06]

MyInformIT

You are here : Home : MyInformIT

Welcome to MyInformIT
Exact Phrase All Words Search Tips

As you find content on InformIT, click the Add to MyInformIT link to add up to twenty-five links to any content on InformIT. [ Account Administration ] [ Logout ] Windows 98 Professional Reference Category: Operating Systems Topic: Windows 98 Introduction From: Using Oracle8 Category: Database Topic: Oracle [ DELETE ]

[ DELETE ]

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising 2000 InformIT. All rights reserved.

http://www.informit.com/My/Accounts [26.05.2000 17:01:18]

InformIT : iBooks

You are here : Home : iBooks

New user? Click here.

Home | Browse | My ibooks | Login | Contact Us

Tell a Friend!

Search Text:

All Books

Search Results

Home New Users Click Here!

q

Open Source: FREE digital books in ibooks.com format!

q q q q

Browse: Coming soon. Promotions: Tell a Friend! ibucks Company Info: About Us Press Room Awards Employment Our Content Providers Include:

immediately access all your books 24x7 over the Internet search across the entire text of every book on the site create a personal digital bookshelf no waiting on deliveries, no shipping costs check out our featured and FREE digital books!

Linux Documentation Project Check out the entire LDP here in our sensible, full-text searchable, easy-to-use format.

Click on any subject category below to register with us and start using your FREE online digital book immediately! > Linux > Java > XML > Windows 2000 Featured Books

The Apache Software Foundation The Apache Software Foundation provides organizational, legal, and financial support for the Apache open-source software projects.

We add titles from new publishers and technology companies every day. Join our content providers!

New User Info ibucks Your Privacy

Our Content Providers Secure Shopping Guarantee Home | Browse | My ibooks | Login | Contact Us MyCart | MyWishList | MyAccount | Help

Employment Press Room About Us

http://www.informit.com/partner/ibooks/ (1 of 2) [26.05.2000 17:01:50]

InformIT : iBooks

Business Phone: 512-478-2700 Copyright 2000 ibooks.com, Inc.

http://www.informit.com/partner/ibooks/ (2 of 2) [26.05.2000 17:01:50]

InformIT : Web Based Training

You are here : Home : Web Based Training

High quality, self-paced training to help achieve your goals conveniently Anytime, Anywhere! No parking hassles, no missed assignments, and nothing to install. Affordable Unlimited Access to over 365 courses. Subscription plans start as low as $48 per person per year.
Demo | Overview | Pricing & Sign-up | Course Login | Course Catalog |Course Features | Tips | Tell a Friend! | Vote

http://www.informit.com/training/ [26.05.2000 17:02:02]

InformIT.com -- Find IT Solutions Here !: Recommend-It To A Friend

Tell a friend about InformIT.com -- Find IT Solutions Here ! You Could Win $10,000 When You Do! Your Name: Friends' Emails: Your E-mail: Include a Message:

Receive our FREE monthly newsletter and periodic notices of hot new Computers & Internet sites. Enter me to win $10,000 and a Sony DVD Player and let me know how I can win other great prizes!
Powered by the FREE Recommend-It Service Learn More Click for our Privacy Policy and Terms of Service.

http://www.recommend-it.com/p.e?415858 [26.05.2000 17:02:08]

MyInformIT

You are here : Home : MyInformIT

MyInformIT
Exact Phrase All Words Search Tips

Content Added Successfully

You have successfully added the following information to your MyInformIT: Using Oracle8 Chapter: Category: Database Topic: Oracle
Return to your original location by hitting your browser's back button or by clicking on the chapter or title link above. If you would like to modify your account information, please use the Account Administration section of MyInformIT.

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising 2000 InformIT. All rights reserved.

http://www.informit.com/My/Add?isbn=0789716534&ti...egory=Database&topic=Oracle&chapter=&elementname= [26.05.2000 17:02:16]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Understanding Oracle8 Backup Options

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q

Types of Failure Archiving Your Database

r r r

Starting Archiving Stopping Archiving The Automatic Archive Process Understanding Cold Backups Understanding Hot Backups Recovery Manager for Windows NT Databases Using Database Exports as a Backup Strategy Understanding Incremental Backups Understanding Standby Databases Understanding Replication Strategies

Backup Options
r r r r r r r

When you're planning a backup strategy, it's useful to consider the day-to-day hazards that eventually cause any database system to fail. No matter how many UPSs or mirrored disks you have, no matter how regulated your computing environment is, every database system will experience an unexpected failure. Database failures can generally be divided into two categories: q Instance failure is generally the result of an Oracle internal exception, operating system failure, or other software-related database failure. Instance failure will be the diagnosis when the necessary Oracle processes (PMON, SMON, DB Writer, Log Writer) are no longer running and the database wasn't shut down normally. Although instance failures can directly or indirectly lead to database corruption, they're generally nondamaging in nature. Often, simply restarting the database allows operations to continue normally. q Media failure, by contrast, is usually far more sinister. Media failure will usually manifest itself by the database being unable to read data it has previously written. Leading causes of media failure include disk drive failure, bad disk block, deleted data files, and damaged file system(s). Media failure, unlike instance failure, almost always results in damage to your database that must be repaired before the database can resume normal operations. Fortunately, Oracle8 provides many methods for recovering from data loss.

Archiving Your Database

Because databases often contain mission-critical information, an organization may not be able to tolerate the

http://www.informit.com/content/0789716534/element_012.shtml (1 of 11) [26.05.2000 17:03:01]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8

loss of any transactions whatsoever. If a database were backed up every night and suffered a disk failure a few minutes before backups were scheduled to begin, you could lose up to around 23 hours' worth of transactions. Oracle provides an elegant solution to this problem in the form of archive logs. Impact of ARCHIVELOG mode on disk requirements Running a database in ARCHIVELOG mode can seriously affect your database's disk needs, depending on how much activity your database has. Because a copy of every write operation is kept, you may need tens or even hundreds of megabytes of disk space to store all the archived redo logs for one day. Oracle keeps a record of most every operation in its redo logs in order to guard against loss of database buffers should the database instance fail. Because these logs, in aggregate, contain everything needed to reconstruct a database from any time in the past, they can be used to recover from media failure. By default, Oracle overwrites the redo log groups in a round-robin fashion. This is sufficient for instance recovery because a log switch forces a checkpoint that, in turn, forces all dirty database buffers to be written to disk. To guard against media failure, however, it's necessary to keep the redo logs archived since at least the last physical database backup (and in practice, you'll want to keep them much longer). Oracle refers to this as running the database in ARCHIVELOG mode; it will archive every redo log file that's filled up. Keep archive logs on disk Although Oracle allows you to spool your archived redo logs directly to tape on many system architectures, you're strongly advised not to do so. Archiving directly to tape is much slower and requires much more effort and testing than does archiving to disk. Disk space is very cheap these days.

Starting Archiving
By default, Oracle doesn't create a database in ARCHIVELOG mode. You have to manually place a database into ARCHIVELOG mode as soon as it's created, but when you do so, Oracle will stay in ARCHIVELOG mode until you return it to NOACHIVELOG mode (at which time your database again becomes more vulnerable to media failure). To check whether a database is running in ARCHIVELOG mode, check the LOG_MODE column in the V$DATABASE table. This is shown in an example: SQL> select * from v$database; NAME CREATED LOG_MODE CHECKPOINT_CHANGE# ARCHIVE_CHANGE# ---- --------- ---------- ------------------ --------------TEST 02/14/98 08:03:01 NOARCHIVELOG 12964 12951 SQL> In this example, the database TEST isn't running in ARCHIVELOG mode, as indicated by NOARCHIVELOG in the LOG_MODE column. Enable ARCHIVELOG mode (general steps) 1. Modify the init.ora file. 2. Shut down the database. 3. Start the database in MOUNT EXCLUSIVE mode. 4. Enable ARCHIVELOG mode. 5. Perform a cold backup. 6. Restart the database normally. Step 1: Modify the init.ora File You must decide a couple of things before editing the init.ora file: q In what directory you will store archive logs q What format you want the filenames of the archived logs to follow Determining what directory to store the archive logs in is very important. If you exhaust all the space available, Oracle will stop virtually all activity until space becomes available again.

http://www.informit.com/content/0789716534/element_012.shtml (2 of 11) [26.05.2000 17:03:01]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8

When Oracle freezes If Oracle suddenly freezes and doesn't respond to the most basic SQL statements, first check to make sure that you haven't used up all the space available to your archive logs. If you move some archive logs to another directory, Oracle will automatically resume database operations. If Oracle must wait for space to become available in the archive log directory, it will log this event in the applicable alert.log file. The following parameters must be set in the database's appropriate init.ora file: q log_archive_start indicates whether Oracle should automatically archive filled redo logs. You should always set this to true. q log_archive_dest is the directory to which you want archive log files to be written. This directory should hold exclusively archived redo logs and have plenty of free space. q log_archive_format defines the format of the archived redo log filenames. Use %s to denote the sequence number of an archived redo log file. A good naming convention is to use the database SID followed by %s (to denote each sequence number), and then an .ARC to denote an archive log. Here's an example of these settings on a UNIX system: log_archive_start = true # if you want automatic archiving log_archive_dest = /opt/oracle803/archive/test log_archive_format = TEST%s.arc A typical Windows NT init.ora setting might look like this: log_archive_start = true log_archive_dest = %ORACLE_HOME%\database\archive log_archive_format = "TEST%S.ARC" Monitoring the archive destination Ideally, you should have a monitoring system in place to constantly keep watch for an archive destination directory that's quickly filling up. By being warned before your destination directory is actually full, you can take corrective measures before database operations are affected. BMC's Patrol product offers this capability, although many other excellent products in the market-place serve this need. Step 2: Shut Down the Database A normal (or immediate) shutdown is required to continue. The following example shows how to shut down the database in a UNIX environment. This example will close any open sessions on the database and roll back any transactions in progress. Who's using the database? By querying the V$SESSION view, you can see who's logged in to the database. V$SESSION can help you identify whether the database is in use and who's using it. oreo:~$ svrmgrl Oracle Server Manager Release 3.0.3.0.0 - Production (c) Copyright 1997, Oracle Corporation. All Rights Reserved. Oracle8 Enterprise Edition Release 8.0.3.0.0 - Production SVRMGR> CONNECT INTERNAL; Connected. SVRMGR> SHUTDOWN IMMEDIATE; Database closed. Database dismounted. ORACLE instance shut down. SVRMGR> Step 3: Start the Database in MOUNT EXCLUSIVE Mode Continuing the example from Step 2, this example shows how to use the STARTUP command with the MOUNT and EXCLUSIVE options:

http://www.informit.com/content/0789716534/element_012.shtml (3 of 11) [26.05.2000 17:03:01]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8

SVRMGR> STARTUP MOUNT EXCLUSIVE; ORACLE instance started. Database mounted. SVRMGR> Step 4: Enable ARCHIVELOG Mode The following ALTER DATABASE ARCHIVELOG command will place the database into ARCHIVELOG mode: SVRMGR> ALTER DATABASE ARCHIVELOG; Statement processed. SVRMGR> Step 5: Perform a Cold Backup Cold backups and changing ARCHIVELOG mode You may want to alternate between ARCHIVELOG and NOARCHIVELOG modes. It's absolutely essential that you perform a cold backup after you re-enable ARCHIVELOG mode. Failure to do so may render your archive logs useless. By following the procedures outlined in Chapter 13, "Selecting and Implementing a Backup Strategy," you must perform a cold backup of the database before continuing. This is necessary because the archived redo logs are useful only when they can be applied to a database backup made since ARCHIVELOG mode was enabled. Step 6: Restart the Database Normally In Server Manager, restart the database normally to allow users back onto the database. The V$DATABASE view will now reflect the switch to ARCHIVELOG mode in the LOG_MODE column.

Stopping Archiving
From time to time, it will be beneficial to stop archiving on your databases. For instance, during a maintenance period, you may be importing or deleting large amounts of data that would generate an excessive number of archive logs. Stopping archiving will still provide recovery in the event of instance failure, but should a media error occur, it will be necessary to restore from the last cold backup. Stop archiving (general steps) 1. Shut down the database. 2. Start the database in MOUNT EXCLUSIVE mode. 3. Enable NOARCHIVELOG mode. 4. Open the database. Step 1: Shut Down the Database Just as when ARCHIVELOG was enabled, the database must first be shut down. The following example shows how to shut down the database in a Windows NT environment. This example will close any open sessions on the database and roll back any transactions in progress: D:\ORANT\BIN\SVRMGR30 Oracle Server Manager Release 3.0.3.0.0 - Production (c) Copyright 1997, Oracle Corporation. All Rights Reserved. Oracle8 Enterprise Edition Release 8.0.3.0.0 - Production SVRMGR> CONNECT INTERNAL; Connected. SVRMGR> SHUTDOWN IMMEDIATE; Database closed. Database dismounted.
http://www.informit.com/content/0789716534/element_012.shtml (4 of 11) [26.05.2000 17:03:01]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8

ORACLE instance shut down. SVRMGR> Step 2: Start the Database in MOUNT EXCLUSIVE Mode The database must be in MOUNT EXCLUSIVE mode to change between NOARCHIVELOG and ARCHIVELOG mode: SVRMGR> STARTUP MOUNT EXCLUSIVE; ORACLE instance started. Database mounted. SVRMGR> Step 3: Enable NOARCHIVELOG Mode Backing up after NOARCHIVELOG Although a cold backup isn't required when switching to NOARCHIVELOG mode, it's recommended that you do so anyway. By making a cold back-up at this time, you'll have a known fallback point to restore to should you experience media failure or data corruption. The ALTER DATABASE NOARCHIVELOG command places the database into NOARCHIVELOG mode: SVRMGR> ALTER DATABASE NOARCHIVELOG; Statement processed. SVRMGR> Step 4: Open the Database Use the ALTER DATABASE OPEN command to open the database for normal activity: SVRMGR> ALTER DATABASE OPEN; Statement processed. At this time, you may want to query the V$DATABASE view to confirm the change in log mode.

The Automatic Archive Process

In the procedure for enabling ARCHIVELOG mode, recall that the following line was added to the init.ora file to allow automatic archiving of the filled redo logs: log_archive_start = true By starting the automatic archiver, the database will copy redo logs to the designated destination directory as each redo log is filled. This asynchronous process won't necessarily begin copying at the instant of a logfile switch, but it usually can keep up with your database's transaction load. Confirming that the archive process is running On UNIX systems, you can confirm that the archive process is running by looking for a process named ora_arch_YourSID.

Backup Options
Document backup and recovery Backup systems often run for quite some time without any DBA intervention required. Unfortunately, memories fade and staff turnover can lead to confusion when the time comes to restore a database. It's essential that backup and recovery procedures be tested and documented. Chances are that during a high stress database recovery, not everyone is going to remember the subtle details that can make or break a recovery effort. The need to back up databases is certainly obvious enough. However, databases-because of their highly structured and transaction-centric nature-have special backup and recovery needs. By carefully reading and understanding the backup strategies Oracle provides, you can implement a reliable backup strategy that meets
http://www.informit.com/content/0789716534/element_012.shtml (5 of 11) [26.05.2000 17:03:01]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8

your organization's needs. The most important aspect of any backup plan is to thoroughly test database restores in a test environment. Backups can often appear to have run properly but be incorrect for recovery situations. It's absolutely imperative that all DBAs have first-hand experience with backup and recovery plans.

Understanding Cold Backups

The simplest and most straightforward backup in Oracle is known as a cold backup. Essentially, a cold backup involves nothing more than shutting down the Oracle database instance and backing up all the relevant database files, including Backing up redo logs Oracle doesn't technically need redo logs backed up. Because the database isn't running, however, they can be safely backed up and will prevent you from having to recreate redo logs should you need to restore the database.
q q q q q

Data files Control files Redo logs Archived redo logs init.ora and config.ora (if applicable)

The key to cold backups is that you must have the database instance shut down before beginning. Although the backup process may very well appear to work with the database running, it's very possible that the backup will be corrupted and unusable. When backing up the database, be sure to also back up all the Oracle program files. All these files are typically found under the ORACLE_HOME directory. This directory tree often contains additional configuration files, such as Net8 files and any applied patches. Unlike many database systems, Oracle doesn't provide a backup and restore system per se. Oracle instead relies on operating system utilities, such as tar in UNIX. Although this may seem to be a weakness at first, it's actually something of a feature. Many organizations spend a great deal of money on complex and robust backup systems far superior to anything any database maker now bundles with their product. Oracle, in keeping with its history of flexibility, lets you use the best backup tools available for your environment to back up the operating system and the database. Oracle's Enterprise Manager Oracle's Enterprise Manager includes a backup and restore utility for Windows NT environments. While this utility is functional, most DBAs find that operating system and third-party backup tools work much better for them. The advantages of cold backups are numerous: q Quick and easy q Fairly trouble-free implementation; most sites simply back up the database files as part of a full system backup q Simple restores q Very little site-specific customization required for implementation The disadvantage of cold backups is that the database must be shut down. If you can afford to shut down a database for backups, cold database backups usually offer the best and easiest backup strategy. Effects of taking down a database Shutting down an Oracle data-base can have lasting effects beyond the actual backup period. When a database is shut down and restarted, the data dictionary cache is blank and there is no data in the database block buffers. The morning data-base activity following a backup cycle could be slowed down, as Oracle must reload the data dictionary and the working set of database blocks into the SGA. SEE ALSO More details on performing a cold backup,

http://www.informit.com/content/0789716534/element_012.shtml (6 of 11) [26.05.2000 17:03:02]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8

Understanding Hot Backups

Oracle uses the term hot backups to describe the process of backing up the database while it's open and available to users. Hot backups are an essential component of 24-hour, 7-days-a-week operations, as it allows for backups necessary to data security without interrupting mission-critical operations. Don't back up redo logs Redo logs aren't backed up during hot backups. In fact, doing so can cause serious database corruption. Hot backups will back up these components in a manner similar to a cold backup system: q Archived redo logs q init.ora and config.ora (if applicable) However, special consideration must be given to data files and control files. The advantage of hot backups is that the database can continue normal operations while the database is being backed up. On the other hand, hot backups have several disadvantages: Testing hot backups The need to test hot backup/recovery systems can't be overemphasized. It's absolutely essential that backups be periodically restored to test machines in order to verify that hot backups are being taken properly and to ensure that the skills needed to restore are kept sharp.
q q q

Much more complex to implement Custom site-specific backup scripts must usually be written Extensive testing required to prove viability

Despite the obvious advantage of hot backups, they typically require considerably more time and effort to successfully implement. While running a 24/7 operation may be the trendy thing to do these days, make sure that your business needs require this availability before incurring the time and expense of hot backups. Although database operations can continue during a hot backup without interruption, it's still important for you to schedule backups during the least amount of database activity (UPDATE, INSERT, and DELETE operations, in particular). Hot backups will cause a database to incur additional overhead in terms of CPU, I/O, and higher production of archived redo logs. If your organization truly needs to run 24/7, hot backups will provide a proven and robust solution to keep your business running and your database safe and secure. SEE ALSO More details on performing a hot backup,

Recovery Manager for Windows NT Databases

Preparing to use Recovery Manager Always make a backup of the damaged database-even if parts of the database are missingbefore using Recovery Manager (or any recovery effort, for that matter). That way you have something to fall back on if anything goes wrong during recovery and the database becomes more badly damaged. An excellent rule of thumb is to back up the database and try to recover it yourself once. If your first attempt fails, restore from the backup you made before trying to restore and call Oracle Worldwide Support for guidance. Oracle8 DBAs supporting Windows NT-based databases have a convenient tool for recovering damaged databases: Recovery Manager (see Figure 12.1). Recovery Manager is a GUI-based tool that simplifies database recovery in these cases: Figure 12.1 : This is the opening window for Oracle's Recovery Manager on Windows NT 4.0.
q q q

Recovering from lost or damaged data file(s) Replacing lost or damaged control file(s) Performing complete restores from a database backup

Recovery Manager also has an automatic recovery option that may be able to automatically recover a database with little or no DBA intervention. This option isn't a silver bullet for solving all database recovery problems, however. Automatic recovery can't work correctly unless the proper up-front work has been done to ensure that
http://www.informit.com/content/0789716534/element_012.shtml (7 of 11) [26.05.2000 17:03:02]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8

database backups are performing regularly and correctly.

Using Database Exports as a Backup Strategy

The backup methods discussed here have been physical in nature. That's to say that the backup methods all copy a database's physical data files verbatim to a backup device. These backups provide a fast backup and are relatively easy to use in a restore process should media failure occur. Backing up just a database's data through Oracle's Export utility is known as a logical backup. In many instances it's useful (and necessary) to use logical backups: q Object or row recovery. If someone inadvertently drops a table or deletes rows from a table, it's exceedingly difficult to restore just a particular table or a few rows from a physical backup. q Major release upgrade of Oracle. When Oracle6 was replaced by Oracle7, Oracle changed the underlying database structures. This necessitated that the data contained in Oracle6 databases be exported and then imported into a new Oracle7 system. Oracle will advise when this operation is necessary. q Migrating an Oracle database between significantly different database servers. The physical data structures between different platforms (MVS, VMS, Windows NT, and UNIX) aren't the same. In order to move data between different Oracle databases on different platforms, it's necessary to export data from the source platform and then import the data into the target system. q Moving data between two instances on the same physical machine. Because a physical backup/restore on the same machine would normally overwrite the source instance, it's necessary to export the data from the source instance and then import it into the target instance. Export is sometimes needed with similar platforms Using exports is advised sometimes even when moving data between machines of the same platform. If the data, redo, and control files won't be placed in the same location, it's advisable to export data from the source machine and then import it into a target system configured to receive the data. Exports can be performed with the database up and running. Oracle's built-in read consistency allows it to have a read-consistent view of each table in the database. However, referential integrity can't be guaranteed because Oracle's read consistency will be applied only on a table-by-table basis. If at all possible, exports should be run with the database running in restricted mode. Regular exports should complement hot or cold backups in any backup strategy. Export is a necessary component primarily to provide recovery from the loss of data due to dropped tables or deleted rows. Hot and cold backups are generally useful only for recovery from media failure. User or application errors are often recoverable only from logical backups.

Understanding Incremental Backups

Incremental exports work at the table level When performing an incremental export, be aware that the full table is exported-not just the changed row(s)-if anything at all changes in a table. Keeping an up-to-date export backup on hand doesn't always require that the complete database be exported. Oracle8 provides an incremental-type export that may reduce the time and space required for an export backup by exporting only tables that have changed since the last full or incremental export. In many database systems, only a few tables are actually updated on a day-to-day basis. Many more tables are relatively static in nature and don't need to be exported during each export cycle to maintain a complete export backup. Many DBAs avoid making export backups regularly because of the time and space required to do so. You may be able to overcome this obstacle by making several incremental export backups between full export backups.

Understanding Standby Databases

Pitfalls with standby databases

http://www.informit.com/content/0789716534/element_012.shtml (8 of 11) [26.05.2000 17:03:02]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8

Standby databases can work only if the delivery of every archive log from the production machine can be guaranteed. If an archive log is lost, it is necessary to resynchronize the stand-by database machine with a fresh hot or cold backup. For this reason, you need to implement an automated delivery system of archive logs from the production database to the standby database. Oracle version 7.7.3 introduced the concept of a standby database, which allows you to configure a database that's close to being up-to-date with an online production database instance. In case of a production instance/machine failure, the standby database can be opened, which allows normal database activity to continue. A standby database is, first, an identical copy of the production database (usually this is done by restoring from a cold backup). From this synchronization point, all archive logs generated by the production database machine will be copied to the standby database machine and applied to the database. The standby database is, essentially, always running in recovery mode because it's actively applying archive logs any time the production database is in operation. Standby databases don't eliminate the need for normal backups on the production database machine. A dropped table or deleted row will also be dropped or deleted on the standby machine. Standby databases have several advantages: q Fairly easy to implement. q Will work with all datatypes on Oracle7 databases. q Most database changes will be copied automatically to the standby database. q Replication has a negligible impact on the production system. Standby databases also have several disadvantages: Standby databases can't be used for load balancing Because the standby database is in recovery mode and not open, it's not available for use by any users. You can't use a standby database to help with load balancing on the production machine.
q q q

Almost never completely up-to-date Can't be used for load balancing Only the entire database can be duplicated; no provision for duplicating just a subset of the production database

Standby databases are usually best suited for disaster-recovery database machines. SEE ALSO Learn how to create a standby database,

Understanding Replication Strategies

Oracle8 provides two technologies for replicating all or part of a database between two or more instances: q Snapshot replication q Symmetric replication Snapshot Replication Limitations in Oracle7 If you're working with some Oracle7 databases, be aware that snapshots can't replicate LONG or LONG RAW datatypes. If a snapshot's base query includes a LONG or LONG RAW column, the result will be NULL values in the target system. All other columns will transfer normally. Snapshot replication produces a copy of database tables on a target instance based on a query to the source database. At the time a snapshot is initially taken, the specified query (maybe an entire table) is run and the resulting data is loaded into the target snapshot table. Oracle provides a fairly sophisticated facility for updating snapshots based on time or update activity. Snapshot replication has the following advantages: q Very easy to implement

http://www.informit.com/content/0789716534/element_012.shtml (9 of 11) [26.05.2000 17:03:02]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8
q q

Excellent for producing a static set of data from OLTP systems to DSS systems Can be used for some limited load balancing

On the other hand, snapshot replication has the following disadvantages: q Snapshots may become out-of-date immediately. q Excessive refreshing can heavily tax system and Internet-working resources. q Updates can occur only on the master database. Updating rows in snapshots It's technically possible to allow updates to a snapshot. However, any changes made won't be sent back to the master database and may be over-written during the next snapshot refresh. Typical applications can include the following: q Transferring OLTP system data to a DSS or data warehousing system for thorough analysis q Transferring data to a dedicated instance to avoid long running batch jobs from adversely affecting the production system q Disaster recovery q Creating a test database environment from production systems SEE ALSO More information on snapshots in relation to setting up a read-only failover database, Symmetric Replication Symmetric replication offers a mission-critical and robust means of keeping two or more instances synchronized. Symmetric replication can ensure that a transaction isn't fully committed until all systems being replicated have committed the transaction locally. Alternatively, it can replicate asynchronously, allowing each database node to run at full speed without holding up local updates because of remote database speed issues. Limitations in Oracle7 If you're working with some Oracle7 databases, be aware that symmetric replication can't replicate LONG or LONG RAW datatypes. Symmetric replication is one of the most complicated units of Oracle and any other relational database. Issues such as network reliability, resolving conflicting updates, and transaction volumes are major design issues that must be planned for and dealt with. The advantages to symmetric replication are as follows: q Updates done on any system can automatically be posted on all other replicated and master systems. q Replicated systems can be configured to be kept completely up-to-date. q It's ideal for load-balancing most systems. Symmetric replication isn't without its disadvantages: q It's more difficult to set up and administer than other replication options. q A high update transaction volume on one machine may stress other systems' resources. q Network outages may bring all update-database activity to a halt. q Potentially very high network resource requirements. Typical applications may include the following: q High availability and disaster-recovery systems q Using many database instances for load-balancing purposes Impact on network resources Symmetric replication will transfer each update transaction from any master database to all other machines that subscribe to database updates. Depending on the volume of updates, this can easily saturate wide are ea networks. Even high-speed local area networks can become bottlenecks during batch update or load cycles. < Back Contents Next >

Save to MyInformIT

http://www.informit.com/content/0789716534/element_012.shtml (10 of 11) [26.05.2000 17:03:02]

informit.com -- Your Brain is Hungry. InformIT - Understanding Oracle8 Backup Options From: Using Oracle8

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_012.shtml (11 of 11) [26.05.2000 17:03:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Performing Database Recovery

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q q

Recovery Strategies Analyzing the Failure and Determining Recovery Options Recovering from the Loss of General Data Files
r r r

Recovering from a Lost Data File in the User Tablespace Recovering from a Lost Data File in a Read-Only Tablespace Recovering from a Lost Data File in an Index Tablespace Database is Down Database Is Up and Running Recovering with a Cold Backup Recovering with a Hot Backup Recovering with Missing Redo Data Recovering with a Mirrored Control File Recovering Without a Mirrored Control File

Recovering from a Lost Data File in a Rollback Tablespace

r r

Recovering from Loss of System Tablespace

r r r

Recovering from the Loss of a Control File

r r

Recovering from Loss of Online Redo Logs

Analyzing the recovery process involves determining the factors that influence the recovery process. Database size, system complexity, database structure, and application structure are the main factors that influence the mean time to recover (MTTR). The MTTR can be very critical to the operation of systems that need high availability. You can reduce the MTTR in several ways: q Reduce the size of components that need to be recovered. q Use Oracle8's table and index partitioning features. Using partitions will minimize the impact of the failure on the rest of the system. q Ensure that the backup can be easily and quickly accessed in the event of a failure. q Test your backups to avoid any surprises. q Ensure that you are familiar with the recovery procedures that have to be followed based on the type of failure. Keep the common recovery scripts handy. q Design the database to promote autonomous components. When the components are autonomous, their impact on the rest of the system is minimal and the recovery is faster.

http://www.informit.com/content/0789716534/element_014.shtml (1 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

Table 14.1 describes the techniques that can be used for high availability. These strategies should be used to quickly recover in the event of a database failure. Table 14.1 High availability strategies Technique Object-level recovery using Export, Import, and SQL*Loader Failover systems using hardware redundancy Oracle standby databases Usage Uses Export/ Import to protect data Failover provided by using another node Primary database's redo log keeps another database updated, which can be used during recovery Uses Oracle's replication feature to provide high availability Advantages Fast object- level recovery No data loss due to redundant system Fast recovery; failover; disaster recovery possible Disadvantages Difficult to scale; you must be aware of object associations No scalability; costly

Oracle Symmetric Replication

Oracle Parallel Server

Triple mirroring EMC SRDF facility

Customized store-andforward replication

Data loss possible; complicated setup and maintenance; potential of replicating data-base corruption No data loss; failover; Slow recovery due to use disaster recovery of transactions; use of possible; both databases two-phase commit can can be used lead to additional simultaneously problems while maintaining the database's consistency Clustering solution that No data loss; fast Tuning can be difficult; allows failover to another failover; protects against application design plays instance; recovery can node and cache failures; significant part in proceed simultaneously high scalability; load strategy's success and is done by the balancing surviving instances Uses a third hardware Fast hot backups; fast Cost of triple writes and that is a mirror recovery resilvering Physical I/O- based No data loss; failover; Potential of replicating replication disaster recovery database corruption possible; faster than Oracle Symmetric Replication Makes use of Oracle8 No data loss; fast Complex; serializing of features such as recovery transactions advanced queuing or trigger-based asynchronous replication

General steps to recover a database system 1. Detect the failure. The detection of an outage is usually simple: Either the database isn't responding to the application, or the system has displayed explicit error messages. However, a problem such as a corrupt control file may not be detected while the database is running. 2. Analyze the failure. You should analyze the type and extent of the failure; the recovery procedure will depend on this analysis. This task can take a significant amount of time in large systems. 3. Determine the components of the database that need recovery. This task can also be significant in large systems. You need to determine which components (such as a table) are lost, and then determine whether you need to recover the tablespace or a data file. 4. Determine the dependencies between components to be recovered. Usually the components aren't isolated; loss or recovery of a database object can affect other objects. For example, if a table needs recovery, you'll also have to recreate the indexes. This step isn't done automatically by the recovery of the table. 5. Determine the location of the backup. The closer the backup is to where the recovery is to be performed, the lesser is the MTTR. Location factors

http://www.informit.com/content/0789716534/element_014.shtml (2 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

If the backup is on a disk, is the disk on-site or off-site? Is the disk local or network? Do you have mirrored copies? Are you recovering from a cold or a hot backup? If the backup is on tape, is the tape on-site or off-site? Do you need additional components to access the tape? 6. Perform the restore. This involves restoring the physical file from disk or tape and placing it at a location where the database can access the file for recovery purposes. The time to restore is affected by file location, file size, file format (raw, export, blocks, or extracts), and possibilities of restore parallelism. 7. Replay redo logs (for archived databases) and resync the database components.

Analyzing the Failure and Determining Recovery Options

You need to answer several questions to determine the type and extent of failure. Your answers dictate the steps you take to recover the system: q What led to the failure: power outage, upgrade, maintenance (OS, hardware, network, database)? q How did the database go down: shutdown abort, crash, normal shutdown? q Are there any operating system errors? q Was the server rebooted? Are there any errors in the operating system log? q Are there any errors in the alert log? q Were trace files generated? q How critical is the lost data? q Are you running Oracle Parallel Server? q Have you attempted any kind of recovery so far? If so, what steps have already been performed? q What's your backup strategy? q If you have cold backups, how was the database shut down when the cold backups were taken? q Are you running in archive log mode? q Do you have all archive logs accessible without any break in the sequence numbers? q Are the online redo logs mirrored? q Are the control files mirrored? q Do you have recent full database export? q What else was running when the system crashed? q Can you bring the instance up? q Can you mount the database? Open it? q What are your system availability requirements? q What size is the database? q Are you using raw files? q What events (if any) are set in your INIT.ORA? q How many rollback segments do you have? Database failures can be detected if you get one of various errors. The following list shows some of the common errors encountered and their solutions: q The message 00205, 00000, "error in identifying controlfile, check alert log for more info" means the system couldn't find a control file of the specified name and size. Check that all control files are online and that they're the same files that the system created at cold start time. q The message 00280, 00000, "change %s for thread %s is in sequence #%s" helps you find the redo log file with the specified change number requested by other messages. Use this information to specify the required archived redo log files for other errors. q The message 00312, 00000, "online log %s thread %s: '%s'" reports the filename for details of another message. Other messages will accompany this message; see the associated messages for the appropriate action to take. q The message 00314, 00000, "log %s of thread %s, expected sequence# %s doesn't match %s" means that the online log is corrupted or is an old version. You need to find and install correct log versions or reset the log.
http://www.informit.com/content/0789716534/element_014.shtml (3 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8
q

The message 00376, 00000, "file %s cannot be read at this time" indicates that Oracle is trying to read from an unreadable file-more than likely, an offline file. Check the file's status and bring it online, if necessary. The message 00604, 00000, "error occurred at recursive SQL level %s" indicates that an error occurred while processing a recursive SQL statement (which applies to internal dictionary tables). Usually this error is accompanied by other errors. If the situation described in the next error can be corrected, do so; otherwise, contact Oracle Support. The message 01110, 00000, "data file %s: '%s'" reports a filename for details of another error. See the accompanying errors for analyzing the problem further. The message 01116, 00000, "error in opening database file %s" usually means that the file isn't accessible. The solution is to restore the database file. The message 01157, 00000, "cannot identify data file %s - file not found" means that the background process couldn't find one of the data files. The database will prohibit access to this file but other files will be unaffected; however, the first instance to open the database will need to access all online data files. Another error from the operating system will describe why the file wasn't found. To solve this problem, make the file available to database and then open the database or do an ALTER SYSTEM CHECK DATAFILES. The message 01194, 00000, "file %s needs more recovery to be consistent" means that an incomplete recovery session was started, but an insufficient number of logs was applied to make the file consistent. The reported file wasn't closed cleanly when it was last opened by the database. You must recover this file to a time when it wasn't being updated. The most likely cause of this error is forgetting to restore the file from a backup before doing incomplete recovery. Either apply more logs until the file is consistent or restore the file from an older backup and repeat the recovery.

Recovering from the Loss of General Data Files

The procedure to recover from the loss of general data files depends on the type of tablespace from which the data file is lost-rollback tablespace, user tablespace, index tablespace, or read-only tablespace. Several symptoms can indicate this problem. You might get the following errors: q You're trying to start the database and get error message ORA-1157, ORA-1110 and possibly an operating system error. q You're trying to shut down the database in normal or immediate mode and get error message ORA-1116, ORA-1110 and possibly an operating system error. These errors can indicate a loss of a general data file. It's up to you to determine the type of tablespace involved.

Recovering from a Lost Data File in the User Tablespace

If you determine that the tablespace contains user data, the steps you follow to recover the tablespace depend on whether you have a good backup and whether that backup was cold or hot. Recovering with a Cold Backup In this case you're in the NOARCHIVELOG mode. The data file recovery will be complete if the redo to be applied is within the range of your online redo logs. Recover with a cold backup 1. Shut down the database. 2. Restore the lost data file from the backup. 3. Start the database. 4. Execute the following query to determine all your online redo log files and their respective sequence and first change numbers: SELECT X.GROUP#, MEMBER, SEQUENCE#, FIRST_CHANGE# FROM V$LOG X, V$LOGILE Y WHERE X.GROUP# = Y.GROUP#; 5. Determine the CHANGE# of the file to be recovered: SELECT FILE#, CHANGE# FROM V$RECOVER_FILE;
http://www.informit.com/content/0789716534/element_014.shtml (4 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

The data file can be recovered if the CHANGE# obtained is greater than the minimum FIRST_CHANGE# of your online redo logs. 6. Recover the data file by using the online redo logs: RECOVER DATAFILE 'fullpath of the datafile' Confirm each log that you're prompted for during the recovery until you receive the message Media Recovery complete. 7. Open the database: ALTER DATABASE OPEN Recovering with a Hot Backup In this case you're in the ARCHIVELOG mode. The data file recovery will be complete if the redo to be applied is within the range of your online logs. Recover with a hot backup 1. Shut down the database. 2. Restore the lost data file from the backup. 3. Start the database. 4. Execute the following query to determine all your online redo log files and their respective sequence and first change numbers: SELECT X.GROUP#, MEMBER, SEQUENCE#, FIRST_CHANGE# FROM V$LOG X, V$LOGILE Y WHERE X.GROUP# = Y.GROUP#; The error that reported the failure should indicate the CHANGE# of the file to recover. If this CHANGE# is less than the minimum FIRST_CHANGE# of your online redo logs, the file can't be completely recovered and you have two choices: r If you can afford losing the database changes since the most recent cold backup, restore the backup and continue with the recovery. r If you can't afford to lose the database changes, you have to recreate the tablespace as described in the following section, "Recovering with Missing Redo Data." 5. Recover the data file by using the archived and the online redo logs: RECOVER DATAFILE 'fullpath of the datafile' Confirm each log that you're prompted for during the recovery until you receive the message Media Recovery complete. Use the online redo logs during recovery If while performing step 5 you're prompted for a non-existing archived log, you need to use the online redo logs to continue with the recovery. Compare the sequence number referenced in the ORA-280 message with the sequence numbers of your online redo logs. Supply the full path name of one of the members of the redo log group whose sequence number matches it. 6. Open the database: ALTER DATABASE OPEN Recovering with Missing Redo Data If redo data is missing, the recovery won't be complete as described in the preceding steps and you'll have to recreate the tablespace. To recreate the tablespace, you can either use a good export script that can easily load the data and recreate the objects in that tablespace or load the data through SQL*Loader. Recover with missing redo data 1. Shut down the database. 2. Mount the database:

http://www.informit.com/content/0789716534/element_014.shtml (5 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

Svrmgrl> Startup mount 3. Offline drop the data file: Svrmgrl> ALTER DATABASE DATAFILE 'fullpath of datafile' OFFLINE DROP; 4. Open the database: Svrmgrl> ALTER DATABASE OPEN; 5. Drop the user tablespace: Svrmgrl> DROP TABLESPACE tablespace_name INCLUDING CONTENTS; 6. Recreate the tablespace and the tablespace objects.

Recovering from a Lost Data File in a Read-Only Tablespace

Because read-only tablespaces are never modified, the recovery solution is very simple: Restore the data file from its last backup to its original location. No media recovery is required in this case. There are two exceptions to this procedure, however: q At the time of the last backup, the tablespace was read-only but it has been made read-write since that time. q At the time of the last backup, the tablespace was read-write but it has been made read-only since that time. In either scenario, recovery can be performed by following the same steps as described earlier in the section "Recovering from a Lost Data File in a User Tablespace," based on the backup type you have.

Recovering from a Lost Data File in an Index Tablespace

If you determine that the tablespace contains user indexes, you should use the following steps, depending on whether you have a good backup and whether it's a cold or hot backup. Recovering with a Cold Backup In this case you're in NOARCHIVELOG mode. The data file recovery will be complete if the redo to be applied is within the range of your online redo logs. To recover from this situation, follow the same steps as described earlier in the "Recovering with a Cold Backup" section of "Recovering from a Lost Data File in a User Tablespace." Recovering with a Hot Backup In this case you're in ARCHIVELOG mode. The data file recovery will be complete if the redo to be applied is within the range of your online logs. To recover from this situation, follow the same steps as described earlier in the "Recovering with a Hot Backup" section of "Recovering from a Lost Data File in a User Tablespace." Recovering with Missing Redo Data If redo data is missing, the recovery won't be complete. You need to recreate the tablespace. To do so, you can either use a script that can be used to easily create the index or manually create the indexes by issuing CREATE INDEX statements. Recover with missing redo data 1. Shut down the database. 2. Mount the database: Svrmgrl> Startup mount 3. Offline drop the data file: Svrmgrl> ALTER DATABASE DATAFILE 'fullpath of datafile' OFFLINE DROP; 4. Open the database:
http://www.informit.com/content/0789716534/element_014.shtml (6 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

Svrmgrl> ALTER DATABASE OPEN; 5. Drop the user tablespace: Svrmgrl> DROP TABLESPACE tablespace_name INCLUDING CONTENTS; 6. Recreate the tablespace and all the previously existing indexes in the tablespace.

Recovering from a Lost Data File in a Rollback Tablespace

This recovery scenario is the most critical one, and you should really work with Oracle Support while performing such a recovery. The main issue involved is that you must make sure the active transactions in the rollback segment aren't lost. In this case, the procedure to follow depends on when the loss was detected.

Database is Down
You're trying to start the database and get ORA-1157, ORA-1110 and operating system errors, as well as determine that the tablespace contains rollback segments. One thing you have to determine is how the database was shut down. Database Was Cleanly Shut Down You're certain that the database was shut down via shutdown normal or shutdown immediate. Check the alert log and look at the last shutdown entry. The following log entry indicates that the shutdown was clean: 'alter database dismount completed: alter database dismount" This may be followed by an attempt you made to start, resulting in the ORA errors and a subsequent SHUTDOWN ABORT by Oracle. Recover a database that has been shut down cleanly 1. In the INITSID.ORA file, change the ROLLBACK_SEGMENTS parameter by removing all the rollback segments in the tablespace to which the lost data file belongs. If you aren't sure of the rollback segments you need to remove, insert a # at the beginning of the line to comment out the entire ROLLBACK_SEGMENTS entry. 2. Mount the database in restricted mode: Svrmgrl> STARTUP RESTRICT MOUNT 3. Offline drop the lost data file: Svrmgrl> ALTER DATABASE DATAFILE 'fullpath of datafile' FFLINE DROP; 4. Open the database: Svrmgrl> ALTER DATABASE OPEN If at this point you receive a message that the statement has been processed, skip to step 7; otherwise, if you get error codes ORA-604, ORA-376, and ORA-1110, continue to step 5. 5. This step should be performed only if the database didn't open in step 5. Shut down the database and edit the INITSID.ORA file as follows: r Comment out the ROLLBACK_SEGMENTS parameter. r Add the following line to list all the rollback segments originally listed in the ROLLBACK_SEGMENTS parameter: _Corrupted_rollback_segments = (rollback1,rollback2,...,rollbackN) Now start the database in restricted mode:

http://www.informit.com/content/0789716534/element_014.shtml (7 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

Svrmgrl> startup restrict mount 6. Drop the rollback tablespace that contained the lost data file: Svrmgrl> drop tablespace tablespace_name including contents; 7. Recreate the rollback tablespace with all its rollback segments and be sure to bring them online. 8. Make the database available for general use: Svrmgrl> alter system disable restricted session; 9. If you had to edit the INITSID.ORA file in step 6, shut down the database and edit the file again as follows: r Uncomment the ROLLBACK_SEGMENTS parameter. r If you had to perform step 6, remove the following line: _Corrupted_rollback_segments = (rollback1,rollback2,...,rollbackN) 10. Start the database. Database Wasn't Cleanly Shut Down In this scenario, the database was shut down, aborted, or crashed. You can't offline or drop the lost data file because it's almost certain that the rollback segments with extents in the lost data file contain active transactions; you must restore the lost data file from backup and apply media recovery. If the database is in NOARCHIVELOG mode, a complete recovery is possible only if the redo to be applied is in the range of your online redo log files. Recover a database that wasn't shut down cleanly 1. Restore the lost data file from a backup. 2. Mount the database. 3. Identify whether the file is offline: Svrmgrl> SELECT FILE#, NAME, STATUS FROM V$DATAFILE; 4. If the file is offline, bring it online: Svrmgrl> ALTER DATABASE DATAFILE 'full path of datafile' ONLINE; 5. Execute the following query to determine all your online redo log files and their respective sequence and first change numbers: SELECT X.GROUP#, MEMBER, SEQUENCE#, FIRST_CHANGE# FROM V$LOG X, V$LOGILE Y WHERE X.GROUP# = Y.GROUP#; 6. The file can't be recovered if the CHANGE# is less than the minimum FIRST_CHANGE# of your online redo logs. You now have two options: r Restore from a full database backup, which may result in data loss. r Force the database to open in an inconsistent state, and then rebuild the database. Open the database in an inconsistent state and rebuild it Be careful here! These steps should be used with extreme caution after taking a full database export-there is a potential for database corruption. 1. Shut down the database. 2. Take a full database backup. 3. Make the following changes in your INITSID.ORA file: r Add the following lines:

http://www.informit.com/content/0789716534/element_014.shtml (8 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

_allow_resetlogs_corruption = true _corrupted_rollback_segments = list of all rollback segments r Comment out the rollback_segments parameter. 4. Do a startup mount. 5. Perform an incomplete recovery of the database: Svrmgrl> RECOVER DATABASE UNTIL CANCEL; 6. When prompted for the file, typeCANCEL. 7. Reset the logs and open the database: Svrmgrl> ALTER DATABASE OPEN RESETLOGS; 8. Rebuild the database by taking a full database export and then importing it to a new database. Rebuilding the database Rebuilding the database is an essential step in this procedure because forcefully opening the database can corrupt the data-base. However, if the CHANGE# is greater than the minimum FIRST_CHANGE# of your redo logs, recover the data file by using the online redo logs: RECOVER DATAFILE 'fullpath of the datafile' Confirm each log that you're prompted for during the recovery until you receive the message Media Recovery complete. 9. Open the database: ALTER DATABASE OPEN

Database Is Up and Running

Don't shut down the database if you've detected a loss of data file in the rollback tablespace while the database is up and running. It's simpler to resolve the situation with the database up than it is with the database down. You can use two approaches to recover. Create new rollback segments You may have to create additional rollback segments in a different tablespace to continue working with the database while the current problem is being addressed. Recover a live database in ARCHIVELOG mode 1. Offline the lost data file: ALTER DATABASE DATAFILE 'fullpath of datafile' OFFLINE; 2. Restore the data file from a backup. 3. Apply media recovery on the data file: RECOVER DATAFILE 'fullpath of datafile'; 4. Bring the data file back online: ALTER DATABASE DATAFILE 'fullpath of datafile' ONLINE; Recover a live database, no matter what mode (slower method) 1. Offline all the rollback segments in the tablespace to which the data file belongs: ALTER ROLLBACK SEGMENT rollback_segment OFFLINE; Repeat this statement for all rollback segments in the affected tablespace. 2. Execute the following query to make sure that the rollback segments are offline before they're dropped: SELECT SEGMENT_NAME, STATUS

http://www.informit.com/content/0789716534/element_014.shtml (9 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

FROM DBA_ROLLBACK_SEGS WHERE TABLESPACE_NAME = 'tablespace_name'; 3. Drop all offline rollback segments by running the following command for each segment: DROP ROLLBACK SEGMENT rollback_segment; 4. If there are rollback segments that you tried to offline, but step 2 shows that they're still online, it means that they have active transactions in them. Run the following query to determine the active transactions: SELECT SEGMENT_NAME, XACTS ACTIVE_TX, V.STATUS FROM V$ROLLSTAT V, DBA_ROLLBACK_SEGS WHERE TABLESPACE_NAME = 'I' AND SEGMENT_ID = USN; If this query returns no rows, all the rollback segments are offline. If this query returns one or more rows with a status of PENDING OFFLINE, check the ACTIVE_TX column for these rollback segments. Segments with a value of 0 will soon go off-line; a non-zero value, however, indicates that you have active transactions that need to be committed or rolled back. Dealing with Active Transactions Execute the following query to identify users who have transactions assigned to the rollback segments: SELECT S.SID, S.SERIAL#, S.USERNAME, R.NAME "ROLLBACK" FROM V$SESSION S, V$TRANSACTION T, V$ROLLNAME R WHERE R.NAME IN ('pending_rollback1','pending_rollback2', ... 'pending_rollbackN') AND S.TADDR = T.ADDR AND T.XIDUSN = R.USN; After you determine which users have active transactions in the "pending offline" rollback segments, you can either ask them to commit or roll back their transaction or you can kill their session by executing the following: ALTER SYSTEM KILL SESSION 'sid, serial#'; The following steps can be performed after you have taken care of the active transactions. Clean up after active transactions 1. Drop the tablespace including contents. 2. Recreate the rollback tablespace. 3. Recreate the rollback segments and bring them online.

Recovering from Loss of System Tablespace

A good backup strategy can be a lifesaver in a situation where a data file from the system tablespace is lost or damaged. In the absence of a good backup, you may be faced with the undesirable alternative of rebuilding the database with possible data loss. The symptoms of the scenario in which the system tablespace needs recovery are as follows: q The database crashed. When trying to start the database, you get ORA-1157, ORA-1110 and possibly an operating system error. q You're trying to shut down the database in normal or immediate mode and get ORA-1116, ORA-1110 and possibly an operating system error. You should use the following methods depending on whether you have a good backup and whether that backup is cold or hot.

Recovering with a Cold Backup

In this case you're in NOARCHIVELOG mode. The data file recovery will be complete if the redo to be applied is within the range of your online logs. Recover with a cold backup

http://www.informit.com/content/0789716534/element_014.shtml (10 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

1. 2. 3. 4.

Shut down the database. Restore the lost data file from the backup. Start the database. Execute the following query to determine all your online redo log files and their respective sequence and first change numbers:

SELECT X.GROUP#, MEMBER, SEQUENCE#, FIRST_CHANGE# FROM V$LOG X, V$LOGILE Y WHERE X.GROUP# = Y.GROUP#; 5. Determine the CHANGE# of the file to be recovered: SELECT FILE#, CHANGE# FROM V$RECOVER_FILE; If the CHANGE# is greater than the minimum FIRST_CHANGE# of your online redo logs, the data file can be recovered by applying the online redo logs. If the CHANGE# obtained is less than the minimum FIRST_CHANGE# of your online redo logs, the file can't be completely recovered, and you have two choices: r If you can afford losing the database changes since the most recent cold backup, restore the backup and continue with the recovery. r If you can't afford to lose the database changes, you have to rebuild the database as described earlier in the section "Recovering with Missing Redo Data." 6. If the CHANGE# is greater than your online redo logs' minimum FIRST_CHANGE#, you can recover the data file by using the online redo logs: RECOVER DATAFILE 'fullpath of the datafile' Confirm each log that you're prompted for during the recovery until you receive the message Media Recovery complete. 7. Open the database: ALTER DATABASE OPEN

Rec covering with a Hot Backup

In this case you are in ARCHIVELOG mode. The data file recovery will be complete if the redo to be applied is within the range of your online redo logs. Recover with a hot backup 1. Shut down the database. 2. Restore the lost data file from the backup. 3. Startup mount the database. 4. Execute the following query to determine all your online redo log files and their respective sequence and first change numbers: SELECT X.GROUP#, MEMBER, SEQUENCE#, FIRST_CHANGE# FROM V$LOG X, V$LOGILE Y WHERE X.GROUP# = Y.GROUP#; 5. Recover the data file by using the archived and the online redo logs: RECOVER DATAFILE 'fullpath of the datafile' Confirm each log that you're prompted for during the recovery until you receive the message Media Recovery complete. Use online redo logs for recovery If you're prompted for a non-existing archived log, you'll need to use the online redo logs to continue with the recovery. Compare the sequence number referenced in the ORA-280 message with the sequence numbers of your online redo logs. Supply the full path name of one of the members of the redo log group whose sequence number matches. 6. Open the database:
http://www.informit.com/content/0789716534/element_014.shtml (11 of 15) [26.05.2000 17:04:02]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

ALTER DATABASE OPEN

Recovering with Missing Redo Data

The recovery won't be complete if redo data is missing. Your option isn't attractive-you'll have to rebuild the database. For this purpose, you can either use the Export/Import utility to take a full database export, or you can do a user/table-level export. SQL*Loader can also be used for this purpose.

Recovering from the Loss of a Control File

The control file is very important for the database. Usually a problem with the control file isn't detected while the database is up and running. If the control file is lost or damaged in such a way that Oracle can't recognize it, a subsequent database startup will result in ORA-205 (error in identifying control file '%s'), along with an operating system-level error. The recovery procedure for this situation depends on whether you have lost one of the control files from a mirrored configuration or all copies of the current control file. The following methods should be used to recover the database based on the particular situation.

Recovering with a Mirrored Control File

In this scenario, you can use the other copies of the control file for getting the database up and running. Recover with a mirrored control file 1. Shut down the instance if it's still running. 2. Find the cause of the loss of control file. Is it due to a hardware problem (disk or controller)? 3. If the hardware isn't a problem, make a good copy of the control file to the location of the lost control file and skip to step 6. 4. If the hardware is the problem, make a good copy of the control file to a reliable location. 5. Edit INITSID.ORA or CONFIGSID.ORA to update the CONTROL_FILES parameter to reflect the control file's new location. 6. Start the database.

Recovering Without a Mirrored Control File

In this case, the recovery steps are a bit more involved. First you have to analyze the situation and determine the answers to the following questions: q Do you have a backup trace of the control file that accurately reflects the current database structure? q If you don't have a backup trace of the control file but the database can mount, proceed with the following steps. In the worst-case scenario, when you've lost all the control files, you have to execute the CREATE CONTROLFILE statement. Recover without a mirrored control file 1. Shut down the instance (if it's not already shut down). 2. Startup mount the database. 3. Get a backup trace of the control file: Svrmgrl> alter database backup controlfile to trace; This command will create a trace file in USER_DUMP_DEST. (To find the value of USER_DUMP_DEST, use the command show parameter USER_DUMP_DEST; at the Svrmgrl> prompt.) Listing 14.1 shows an example of such a trace file. Modify the trace file by removing the header information (lines 1-21). Also make any changes desired (such as MAXLOGFILES on line 30 and MAXDATAFILES on line 32) and save the file as create_control.sql. Do a shutdown normal for the database. Take a full backup of the database to protect from any further problems. Execute the following to create a new control file:

4.

5. 6. 7.

http://www.informit.com/content/0789716534/element_014.shtml (12 of 15) [26.05.2000 17:04:03]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

Svrmgrl> @create_control.sql 8. Open the database by executing the following at the server manager prompt: svrmgrl> Alter database open; 9. Shut down the database. 10. Take a full database backup. Listing 14.1 Example control file creation script 01: 02: 03: 04: 05: 06: 07: 08: 09: 10: 11: 12: 13: 14: 15: 16: 17: 18: 19: 20: 21: 22: 23: 24: 25: 26: 27: 28: 29: 30: 31: 32: 33: 34: 35: 36: 37: 38: 39: 40: 41: 42: 43: 44: 45: 46: 47: 48: 49: 50: 51: Dump file E:\ORANT\rdbms80\trace\ORA00167.TRC Tue Mar 31 17:06:56 1998 ORACLE V8.0.3.0.0 - Production vsnsta=0 vsnsql=c vsnxtr=3 Windows NT V4.0, OS V5.101, CPU type 586 Oracle8 Enterprise Edition Release 8.0.3.0.0 - Production With the Partitioning and Objects options PL/SQL Release 8.0.3.0.0 - Production Windows NT V4.0, OS V5.101, CPU type 586 Instance name: sjr Redo thread mounted by this instance: 1 Oracle process number: 8 pid: a7 Tue Mar 31 17:06:56 1998 Tue Mar 31 17:06:56 1998 *** SESSION ID:(7.1) 1998.03.31.17.06.56.062 # The following commands will create a new control file # and use it to open the database. # Data used by the recovery manager will be lost. # Additional logs may be required for media recovery of # offline data files. Use this only if the current # version of all online logs are available. STARTUP NOMOUNT CREATE CONTROLFILE REUSE DATABASE "SJR" NORESETLOGS NOARCHIVELOG MAXLOGFILES 32 MAXLOGMEMBERS 2 MAXDATAFILES 254 MAXINSTANCES 1 MAXLOGHISTORY 899 LOGFILE GROUP 1 'E:\ORANT\DATABASE\LOGSJR1.ORA' SIZE 200K, GROUP 2 'E:\ORANT\DATABASE\LOGSJR2.ORA' SIZE 200K DATAFILE 'E:\ORANT\DATABASE\SYS1SJR.ORA', 'E:\ORANT\DATABASE\RBS1SJR.ORA', 'E:\ORANT\DATABASE\USR1SJR.ORA', 'E:\ORANT\DATABASE\TMP1SJR.ORA', 'E:\ORANT\DATABASE\INDX1SJR.ORA' ; # Recovery is required if any of the datafiles are # restored backups, or if the last shutdown was not # normal or immediate. RECOVER DATABASE # Database can now be opened normally.

http://www.informit.com/content/0789716534/element_014.shtml (13 of 15) [26.05.2000 17:04:03]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

52:

ALTER DATABASE OPEN;

Recover without an accurate trace file 1. Shut down the database. 2. Take a full database backup, including all the data files and redo log files. 3. Use Server Manager and do a STARTUP NOMOUNT of the database. 4. Issue a CREATE CONTROLFILE statement such as the following: Full syntax available For the CREATE CONTROLFILE statement's complete syntax, see Oracle's SQL reference manual. Create Controlfile reuse database "TEST" noresetlogs noarchivelog Maxlogfiles 50 Maxlogmembers 3 Maxdatafiles 500 Maxinstances 8 Maxloghistory 500 Logfile Group 1 '/u01/oracle/8.0.4/dbs/log1test.dbf' size 1M, Group 2 '/u01/oracle/8.0.4/dbs/log2test.dbf' size 1M, Group 3 '/u01/oracle/8.0.4/dbs/log3test.dbf' size 1M, Datafile '/u01/oracle/8.0.4/dbs/systest.dbf' size 40M, '/u01/oracle/8.0.4/dbs/data1test.dbf' size 10M, '/u01/oracle/8.0.4/dbs/data2test.dbf' size 20M; 5. Perform media recovery on the database: Svrmgrl> Recover database; 6. Open the database: Svrmgrl> alter database open; 7. Do a shutdown normal of the database. 8. Take a cold backup of the database.

Recovering from Loss of Online Redo Logs

Redo logs are important for performing up-to-the-minute database recovery. Several symptoms and errors can indicate that the redo logs are lost or corrupted: q The database crashed, and you get ORA-1194 upon trying to start the database. q You're performing media recovery with the following command and get ORA-314, ORA-312, and ORA-376: "RECOVER DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL" These errors indicate that the redo logs are misplaced, removed, corrupt, or from an old backup. The recovery technique that you need to use depends on whether you've mirrored your redo logs. Recover with mirrored redo logs 1. Identify the files that need recovery: Svrmgrl> select * from V$RECOVERY_FILE; Shut down the database. Bring all the redo log files to a constant state by copying the archive log with the same sequence number as the redo log that was lost to the original redo log. Start the database. If you've lost the current redo logs and your archive logs, the only way that you can open the database is

2. 3. 4. 5.

http://www.informit.com/content/0789716534/element_014.shtml (14 of 15) [26.05.2000 17:04:03]

informit.com -- Your Brain is Hungry. InformIT - Performing Database Recovery From: Using Oracle8

by forcing it to open. This involves corrupting the database. Recover without mirrored redo logs 1. Shut down the database. 2. Take a full database backup. Caution: Database can become corrupt if you force it to open These steps use some very dangerous parameters that should be used only on understanding their consequences and with the help of an Oracle Support Services analyst. After the data-base is opened in this manner, you should rebuild the data-base at your earliest chance. 3. Make the following changes in your INITSID.ORA file: r Add the following lines: _allow_resetlogs_corruption = true _corrupted_rollback_segments = list of all rollback segments r Comment out the rollback_segments parameter. 4. Perform a startup mount. 5. Perform an incomplete recovery of the database: Svrmgrl> RECOVER DATABASE UNTIL CANCEL; 6. When prompted for the file, typeCANCEL. 7. Reset the logs and open the database: Svrmgrl> ALTER DATABASE OPEN RESETLOGS; 8. Rebuild the database by taking a full database export and then importing it to a new database. Rebuild the database after forcing it to open Rebuilding the database is an essential step in this procedure because a forced database-open can corrupt the database. The following parameters can have a very harmful effect on the database and should be used carefully: q _allow_resetlogs_corruption When this parameter is set to TRUE, it will allow resetlogs even if there are hot backups that need more redo applied or the data files are out of sync for some other reason. It's effective only if you open the database with the RESETLOGS option: ALTER DATABASE OPEN RESETLOGS; _corrupted_rollback_segments If a rollback segment isn't accessible because the file it's in is corrupted or offline, you can force the system to come up without the rollback segment by specifying the segment in this parameter. This parameter prevents the rollback of active transactions in the specified corrupted rollback segments. _offline_rollback_segments This parameter prevents the rollback of active transactions in the listed offline rollback segments. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_014.shtml (15 of 15) [26.05.2000 17:04:03]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Using Recovery Manager for Backup and Recovery

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

Defining Recovery Manager

r r r r r r r

Backing Up the Whole Database or Specific Parts Backup Sets and Image Copies Stored Scripts Parallel Operations Recovery Manager Reports Corruption Detection System Performance Creating a Database Schema for the Recovery Catalog Creating the Recovery Catalog Recovery Manager Backup Features Recovery Manager Scripting Commands Executing a Backup Script The ALLOCATE CHANNEL Command RELEASE CHANNEL SETLIMIT CHANNEL BACKUP The COPY Command Restoring the Full Backup More on Restores

Setting Up the Recovery Catalog

r r

Implementing Your Backup Strategy

r r r

rman Commands
r r r r r

Restores
r r

q q q q

Using the REPORT and LIST Commands Learn about Recovery Manager, the newest Oracle utility for ensuring secure restorable backups Get Recovery Manager up and running See Recovery Manager in action, performing a full backup and full restore

Defining Recovery Manager

The most important responsibility for any computer system manager is maintaining a good, solid backup
http://www.informit.com/content/0789716534/element_015.shtml (1 of 16) [26.05.2000 17:05:19]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

procedure. Whether you manage a large data center with many systems or just your own PC, you need to be prepared in case of a system failure. Recovery Manager is the utility provided with the Oracle Server software that's used to perform database backups and restores. Recovery Manager is supported under Oracle8 and later and replaces the Enterprise Backup Utility (EBU) provided with Oracle7. It offers more than traditional cold backups via the operating system-it even offers online backups with the tablespaces in hot backup mode. Recovery Manager has two user interfaces; this chapter focuses on the command-line interface, which you access through the rman utility. Most of you will use Recovery Manager and the recovery catalog to back up multiple databases, and will want to automate those operations. You can use the Backup Manager if you prefer a GUI-based interface; it comes with the Oracle Enterprise Manager software shown in Figure 15.1 (for those of you running Microsoft Windows on some of your client PCs). The OEM is covered in Chapter 4 "Managing with Oracle Enterprise Manager (OEM)." Figure 15.1 : Oracle Enterprise Manager is your entry point to the GUI version of Recovery Manager. Recovery Manager becomes an indispensable tool when used with a recovery catalog. A recovery catalog takes the form of an Oracle schema stored in a database separate from the databases you're backing up; it maintains all information relevant to the structure and backup history of the databases it backs up. What this means is that in case of a system failure, Recovery Manager can handle all tasks relevant to getting the database back up and running. Recovery Manager can perform backups, store backup and restore scripts that can be executed repeatedly, and offer a wide range of options for backing up your databases. Recovery Manager will back up and restore Oracle databases that are running Oracle Server version 8.0 and later. You should continue using Oracle Enterprise Backup Utility (EBU) or whatever homegrown backup procedures you have for Oracle7 databases.

Backing Up the Whole Database or Specific Parts

You can back up your database in several different ways via Recovery Manager. By specifying a full backup, Recovery Manager will back up all your data files or all your archive logs, but not both. You can write the backup to disk or tape. Full backup workaround There's one quick way around the full backup restriction of not being able to back up the archived logs with the data files: Perform a full backup of the archive logs right after the back-up of the data files within the same backup script. When performing a full backup, Recovery Manager skips over database blocks that have never been used, thus speeding the backup operation. A full backup doesn't reset the individual database blocks' "backup flag." The next incremental backup will back up all database blocks modified since the previous incremental backup. The full backup has no impact on what's backed up during an incremental backup. You can perform incremental backups on the data file, tablespace, or database level. During an incremental backup, Recovery Manager backs up all database blocks that have been modified since the last incremental backup. Again, database blocks that have never been used are skipped over.

Backup Sets and Image Copies

Recovery Manager supports two different output formats for backup and restore operations: Sizing backup sets Because multiple database files are written into a single backup set, you run the risk of exceeding your particular operating system platform's maximum file size. For example, if you're using Sun Solaris 2.51, you have a 2GB restriction on file size. Recovery Manager allows you to specify how many files to put in a backup set and how large the back-up set file can become. If the back-up set reaches its maximum size, the file is closed and a new backup set file is created. By tweaking Recovery Manager backup parameters a little, you shouldn't have any problems with file-system restrictions.
q

A backup set is a single file that contains one or more backed-up database file. The backup is performed serially and may be done to disk or tape. An image copy is just what you imagine it is-it's essentially a second copy of whatever database file

http://www.informit.com/content/0789716534/element_015.shtml (2 of 16) [26.05.2000 17:05:19]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

you're backing up. It's the same as having the database down and using the operating system copy command for your backups. Image file copies can't be used during tape backup operations.

Stored Scripts
Recovery Manager lets you define backup operations in the form of scripts. Those scripts can be stored in disk files or can be loaded into the recovery catalog in much the same way a PL/SQL script is stored in the Oracle server. By using stored scripts, you reduce the possibility of operators performing backups introducing errors. You must have the recovery catalog installed to use stored scripts. You learn about how to create a stored script later in the "Recovery Manager Scripting Commands" section.

Parallel Operations
Recovery Manager can run backups and restores in parallel. If your system has more than one tape drive, Recovery Manager can use both of them at the same time, cutting the elapsed time required for the backup. This will work for disk-to-disk and disk-to-tape backup operations. How the parallelization is done is automatic, and is set on or off in the backup command you use. One sample backup later in this chapter uses parallelization.

Recovery Manager Reports

You use two commands with Recovery Manager to generate reports about the backups you've done and the database objects that need to be backed up: REPORT and LIST. They can give you information on the following: q Which database files haven't been backed up lately and need backing up q Which backup sets are obsolete q Which files are no longer recoverable q Which database files are in which backup set You learn how to generate reports from Recovery Manager later in the chapter.

Corruption Detection
The Oracle Server process doing the backup will detect corrupt database blocks during the backup operation and records any corruption in the control file and the alert log. Recovery Manager reads this information after the backup operation completes and stores it in the recovery catalog. Not all types of corruption are detectable at this time, though.

System Performance
With the previous mention of parallelization, you might be thinking, "If I let Recovery Manager operate in this mode, how will I keep it from using all the system's resources?" You can throttle Recovery Manager with the use of the channel control commands. You use the channel control commands to specify limits on disk I/O during backup operations, determine how many threads will be executing concurrently during a parallel operation, and specify the maximum size for the backup pieces you're creating. By using the channel control commands effectively, you can have your backup operations running quickly and efficiently, without affecting the interactive users that may be using the system during your backup.

Setting Up the Recovery Catalog

Using a recovery catalog with Recovery Manager is optional. Recovery Manager becomes much more powerful when a recovery catalog is implemented, which makes database backups and restores much easier. Recovery Manager flexibility Recovery Manager gives you a great many options for implementing your database backup strategy. In this chapter you learn about all the software's capabilities, although in some installations you wouldn't use all of Recovery Manager's capabilities. As mentioned earlier, the recovery catalog takes the form of an Oracle user schema. We will divide the tasks for creating the catalog into two sections. The first section covers the creation of the database schema that will contain the catalog. The second section will handle the steps necessary to create the recovery catalog within the schema.

http://www.informit.com/content/0789716534/element_015.shtml (3 of 16) [26.05.2000 17:05:19]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

Creating a Database Schema for the Recovery Catalog

Let's start by making a few quick decisions regarding the catalog and the database schema that we're creating. In this chapter, let's assume that you have many databases to back up, and that your systems must be up 24 hours a day, 7 days a week. Let's also assume that when a system does go down, it needs to be brought back online as quickly as possible. Recovery catalog default tablespace The recovery catalog database schema should have its own table-space; it should not have its objects stored in the default user table-space. For the examples in this chapter, the recovery catalog will be created in a separate database whose SID is rcover. The schema that the catalog will reside in will be called recman, and the database that we will be backing up has a SID of jdbase. (Of course, you can and will want to use your own names for these objects.) Create the Default Tablespace for the Recovery Catalog Schema Before you can create an Oracle user schema with its own default tablespace, you first need to have the tablespace created. Oracle recommends that the typical recovery catalog require 100MB of storage for one year. That 100MB value is what we will use to create the tablespace that will contain our catalog. You create the tablespace from within SQL*Plus by using the command CREATE TABLESPACE. The following code shows how the RECOVER tablespace was created on my test system: Keep the catalog and target database separate The database that contains the recovery catalog should never be on the same logical disk drive as the database you're backing up, including RAID and mirrored disk drives. The optimal location for the database containing the recovery catalog is a completely separate system. SQL*Plus: Release 8.0.4.0.0 - Production on Thu Apr 23 19:44:49 1998 (c) Copyright 1997 Oracle Corporation. All rights reserved.

Connected to: Oracle8 Enterprise Edition Release 8.0.4.0.0 - Production PL/SQL Release 8.0.4.0.0 - Production SQL> 2 3 4 create tablespace recover datafile '/ora02/oradata/rcover/recover01.dbf' default storage (initial 1m next 1m pctincrease 0) size 100m;

Tablespace created. Now, just to double-check that the table is there and created correctly, use a SELECT command to query the DBA_TABLESPACES view and look up the new tablespace: SQL> select tablespace_name, status, contents, logging 2 from dba_tablespaces 3 where tablespace_name='RECOVER'; TABLESPACE_NAME STATUS CONTENTS LOGGING ------------------------------ --------- --------- -------RECOVER ONLINE PERMANENT LOGGING With the new recover tablespace created and online, you're now ready to create the schema account that will store the recovery catalog. Picking an authentication method

http://www.informit.com/content/0789716534/element_015.shtml (4 of 16) [26.05.2000 17:05:19]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

A good rule of thumb for deciding whether to use passwords or external authentication for schema accounts is to use passwords for any accounts being accessed from anywhere other than the local system. For data-bases that will be accessed from the local system itself, you can feel safe in allowing external authentication. Create the Database Schema You can also create the new user schema for the recovery catalog from within SQL*Plus. Let's create the new account with a password for security reasons. Remember to set the default tablespace to be our RECOVER tablespace. Don't forget to grant the RECOVERY_CATALOG_OWNER role to the schema account, as specified in the Oracle documentation. The SQL*Plus CREATE USER command used to create our schema is as follows: SQL> create user recman identified by recman 2 default tablespace recover 3 temporary tablespace temp; User created. SQL> grant recovery_catalog_owner to recman; Grant succeeded. Again, you'll want to query the DBA tables to make sure that the account was successfully created. The following code verifies that the default tablespace is correctly set and that the account has a password: SQL> select username,password,default_tablespace 2 from dba_users 3 where username = 'RECMAN'; USERNAME PASSWORD DEFAULT_TABLESPACE ---------------- ------------------- ------------------RECMAN 37234A26A0BB0E9F RECOVER With your database schema created, you can now move on to the next section and create the recovery catalog in the recman schema.

Creating the Recovery Catalog

The creation of the actual catalog is fairly simple after you create the catalog tablespace and schema account. All you need to do is run the catrman script, which is located in the rdbms/admin directory (subdirectory of ORACLE_HOME). Use the right schema You must execute catrman from within the RECMAN database schema. Don't execute it from the SYS schema or your own personal schema; otherwise, you'll see errors when performing Recovery Manager operations. You install the catalog by connecting to the database with the RECMANschema and executing the script: ash$ sqlplus SQL*Plus: Release 8.0.4.0.0 - Production on Sat Apr 25 18:37:57 1998 (c) Copyright 1997 Oracle Corporation. Enter user-name: recman Enter password: Connected to: Oracle8 Enterprise Edition Release 8.0.4.0.0 - Production All rights reserved.

http://www.informit.com/content/0789716534/element_015.shtml (5 of 16) [26.05.2000 17:05:19]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

PL/SQL Release 8.0.4.0.0 - Production SQL> @?/rdbms/admin/catrman The script takes a few minutes. When it's done, you should select the table names from the DBA_TABLES view to verify that they are there, in the correct schema: SQL> select table_name from user_tables; TABLE_NAME -----------------------------AL BCB BCF BDF BP BRL BS CCB CCF CDF CKP TABLE_NAME -----------------------------DB DBINC DF DFATT OFFR ORL RCVER RLH RR RT SCR TABLE_NAME -----------------------------SCRL TS TSATT 25 rows selected.

Implementing Your Backup Strategy

Recovery Manager provides many different options for backing up and restoring database files. It can perform full and incremental backups, and can use disks and tapes as backup devices. If you have multiple tape drives on your system, Recovery Manager can take advantage of that and run multiple backups concurrently.

Recovery Manager Backup Features

When you read about the creation of the recovery catalog earlier in this chapter, it was mentioned that the catalog is an optional feature and that Recovery Manager can run without it. Let's take a deeper look at this, and the effect that running without a recovery catalog has on the restorability of your database files. Using the Recovery Catalog with Recovery Manager As you already know, all information regarding the structure, files, and condition of the database is stored in the database's control file. Recovery Manager reads this information out of the control file and updates the recovery catalog. The maintenance of the information in the recovery catalog isn't dynamic. You update the data in the recovery catalog with the rman command RESYNC CATALOG.
http://www.informit.com/content/0789716534/element_015.shtml (6 of 16) [26.05.2000 17:05:19]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

How often you need to perform the resync depends on how active your database is. If your database is in ARCHIVELOG mode and is being updated regularly, you'll want to resync the catalog every few minutes or hours. If your database is static, a greater elapsed time period would probably be sufficient. As an absolute minimum, your resync should be more often than the CONTROL_FILE_RECORD_KEEP_TIME setting in the target database. Catalog sync frequency You want to err on the side of syncing the control file and recovery catalog too often. If the recovery catalog isn't current at the time of the system failure, you have to catalog all backups and changes between the last resync and the system failure to bring the database back up. When the recovery catalog is fully in sync with the control file, a database restoration after a system failure is very straightforward. Not Using the Recovery Catalog Because all the critical data regarding a database's files and structure are stored in the control file, you don't necessarily need to use the recovery catalog when using Recovery Manager. You can simply do your backups and restores the way you always have, and protect the control file to ensure a successful restore. Do both if it's possible Use a catalog with Recovery Manager and continue to maintain your redundant copies of the control file. It's cheap insurance against a prolonged down-time. Without a recovery catalog, however, the following Recovery Manager features will be unavailable to you: q A point-in-time tablespace recovery. q Storing any of your backup scripts within Recovery Manager. q Recovery Manager can't perform restore and recovery operations when your target database's control file is corrupted or deleted. Also, if you decide not to use a recovery catalog, you have to protect the control file the same way you had to under Oracle7-using multiple copies, frequently performing backups, and keeping a reliable record system of when those backups were done.

Recovery Manager Scripting Commands

Before you can perform any backups with Recovery Manager, you first need to look at how to script the backups. The command-line interface to Recovery Manager is called rman, and all functions relating to database backups and restores are executed through this utility. You can perform backup and restore operations when you're running rman and are connected to the recovery catalog and the target database. You use rman in very much the same way you use SQL*Plus-you can execute commands from a script file, from a stored script, or by entering it interactively. Look at Listing 15.1. It's a short script that will add a full backup of a database-to-disk script to the recovery catalog. Listing 15.1 FULLBACK.RCV-Recovery Manager script to create a full backup stored script 01: 02: 03: 04: 05: 06: 07: 08: 09: 10: 11: 12: replace script fullback { # # full backup of database files, excluding archived logs # allocate channel d1 type disk; backup incremental level 0 tag fullback filesperset 50 format '/ora03/backup/%d/%d__t%t_s%s_p%p' (database); }

This simple script performs a full backup of database files, excluding archived logs. You can perform this
http://www.informit.com/content/0789716534/element_015.shtml (7 of 16) [26.05.2000 17:05:19]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

backup interactively if you replace the statement replace script fullback on line 1 with the command run. The following code shows this script's output. It doesn't execute the backup-it simply stores the script in the recovery catalog as a stored backup script. $ rman target jduer/baseball rcvcat recman/recman@rcover cmdfile fullback.rcv

Recovery Manager: Release 8.0.4.0.0 - Production RMAN-06005: connected to target database: JDBASE RMAN-06008: connected to recovery catalog database RMAN> replace script fullback { 2> # 3> # full backup of database files, excluding archived logs 4> # 5> allocate channel d1 type disk; 6> backup 7> incremental level 0 8> tag fullback 9> filesperset 50 10> format '/ora03/backup/%d/%d__t%t_s%s_p%p' 11> (database); 12> } 13> RMAN-03022: compiling command: replace script RMAN-03023: executing command: replace script RMAN-08086: replaced script fullback Recovery Manager complete.

Executing a Backup Script

With the sample backup script successfully stored in the recovery catalog, it's time to perform your first backup with Recovery Manager. To execute a stored Recovery Manager script, you first use the rman utility to connect to the recovery catalog and the target database. When connected, you execute the stored script via the rman run command, as follows: ash$ rman rcvcat recman/recman@rcover

Recovery Manager: Release 8.0.4.0.0 - Production RMAN-06008: connected to recovery catalog database

RMAN-03022: RMAN-03023: RMAN-08030: RMAN-08500: RMAN-03022: RMAN-03023: RMAN-08008: RMAN-08502: RMAN-08010: RMAN-08011:

compiling command: executing command: allocated channel: channel d1: sid=11

allocate allocate d1 devtype=DISK

compiling command: backup executing command: backup channel d1: starting datafile backupset set_count=3 set_stamp=331940454 channel d1: including datafile 1 in backupset channel d1: including current controlfile in

http://www.informit.com/content/0789716534/element_015.shtml (8 of 16) [26.05.2000 17:05:20]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

backupset RMAN-08010: channel d1: including datafile 5 in backupset RMAN-08010: channel d1: including datafile 3 in backupset RMAN-08010: channel d1: including datafile 2 in backupset RMAN-08010: channel d1: including datafile 4 in backupset RMAN-08013: channel d1: piece 1 created RMAN-08503: piece handle= /ora03/backup/JDBASE/JDBASE__t331940454_s3_p1 comment=NONE RMAN-03023: executing command: partial resync RMAN-08003: starting partial resync of recovery catalog RMAN-08005: partial resync complete RMAN-08031: released channel: d1 Connects to the database you're going to back up EXECUTE command is encapsulated inside the RUN command; it doesn't work on its own

rman Commands
In this part of the chapter, you take a look at the syntax of the main rman commands for backup operations-you look at the channel control commands ALLOCATE CHANNEL and RELEASE CHANNEL. The section follows up with details on the two commands for executing backups: backup and copy.

The ALLOCATE CHANNEL Command

The ALLOCATE CHANNEL command establishes a link between Recovery Manager and the target database. The easiest way to visualize an rman channel is to think of it as a thread. You can use multiple channels within your backup script, and Recovery Manager will try to use as many of those channels concurrently as possible. There are some restrictions, though. A single channel can work on only one image file copy or backup set at a time. If you break your backup into two backup sets, however, and create two channels (one pointing to each tape drive), the parallelization of Recovery Manager kicks in and the software will create both backup sets concurrently. For disk-to-disk backups, you can create as many channels as you want. For system performance's sake, however, you'll want to spread the backup set destinations over several different disk drives. This way you don't create I/O contention by having the channels compete for system resources. (We will discuss an example of how to use the parallelization of Recovery Manager later in the section "The COPY Command.") Only one operand is specified with the ALLOCATE CHANNEL command-a name for the channel, such as D1, for the first disk channel. This name comes into play when using the RELEASE CHANNEL command-you use the name to specify which allocated channel to release. Here is a summary of the command parameters the ALLOCATE CHANNEL command accepts: q connect You use this parameter to specify a database connect string for Recovery Manager to use when connecting to the target database. Oracle notes that this parameter is intended to give databases that are running the Parallel Server option the capability to execute the database backup across different instances of the database, on different clustered systems. Using this capability greatly increases throughput to the backup output devices and reduces resource contention of the systems performing the backup. q format You use this parameter to specify a default filenaming convention for the backup sets created by Recovery Manager on this particular channel. The format specifiers are very similar to those used in the init.ora file for naming archived log files. Using the format parameter in your backup command overrides this default. q name This parameter is used to specify the device name of the system hardware that will be performing the backup or restore. The values for this parameter are port-specific, but in many cases are the physical name of a tape drive. Requirements for name parameter The name parameter is only for sequential I/O devices, such as tape drives. You don't use this parameter when opening a channel to disk.
q

parms This parameter is also port-specific, and is a character string that specifies information about the system device you're allocating. Again, this parameter isn't used when allocating a disk.

http://www.informit.com/content/0789716534/element_015.shtml (9 of 16) [26.05.2000 17:05:20]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8
q

type This operand specifies your backup device's device type. You can specify this type to be "disk", or a quoted character string specifying some type of platform-specific device (usually "tape"). See your platform-specific Oracle documentation for more information on your particular system.

The following sample code shows how to allocate a disk channel named d1: allocate channel d1 type disk;

RELEASE CHANNEL
This command is used to deallocate a channel you created with the ALLOCATE CHANNEL command. This command takes only one operand-the name of the channel to release-which you specified during the ALLOCATE CHANNEL command. The following code line shows a DEALLOCATE command's syntax: deallocate channel d1; Freeing system devices You'll want to use this command if your operating system platform actually allocates resources on the system level during the ALLOCATE CHANNEL command. This way you'll have the system device allocated to you during the time that you need it; you can release this device for others to use during times when other devices are being used.

SETLIMIT CHANNEL
You use the SETLIMIT CHANNEL command to throttle the use of system resources for a particular channel. This way you can restrict throughput on a per-channel basis. This command takes one parameter, the channel name, and uses any or all of the following three parameters: q readrate This parameter lets you control the read I/O rate for the channel. An I/O rate is calculated by multiplying your DB_BLOCK_SIZE and DB_FILE_DIRECT_IO_COUNT. This parameter is specified in I/Os per second. You use this parameter to keep your backups from hammering your disk drives and causing disk contention with other users. q kbytes You use this parameter to specify the maximum size of the output files being created during a backup operation. This helps when your backup sets become large and are in danger of exceeding the operating system's largest possible file size. If this threshold is reached when creating backup sets, the current file is closed and a new one is created. q maxopenfiles This parameter controls how many files can be opened by the channel at any particular time. This parameter is useful in preventing the backup from exceeding a system- or process-level quota on the number of files that can be open. Controlling file access If you don't specify the maxopenfiles parameter and don't use the set limit channel command at all, a default of 32 maximum open files is used. The following sample command shows how you would set the channel d1 to open only 48 files at any time, and restrict the read I/O rate to 256: setlimit channel d1 maxopenfiles 48 readrate 256;

BACKUP
As discussed earlier, there are essentially two types of backups: full and incremental. The default, a full backup, selects all database files with the exception of archived logs. In an incremental backup, only database blocks that have been modified since the last backup are written to the backup set. Database blocks that have never been used aren't written out to the backup set, regardless of whether the backup is full or incremental. Backup Object List In your BACKUP command, you use a backup object list to specify which database components you want written into the backup set. There are eight possible values for this backup command operand: Understanding backup levels

http://www.informit.com/content/0789716534/element_015.shtml (10 of 16) [26.05.2000 17:05:20]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

Incremental backups are multilevel, with incremental level 0 a backup of all database blocks. Incremental level 0 is essentially the same as a full backup, except a full backup doesn't affect subsequent incremental backups. This means that if you want to perform a complete backup followed by a series of incremental backups, you'll want to perform an incremental backup level 0 for your complete (full) backup and use incremental level 1 for your nightly incremental backups.
q

database All database data files, including the database control file, are written into the backup set when you specify a value of database. tablespace A value of tablespace means that you want one or more database tablespaces to be included in the backup set. After the keyword tablespace, you follow with a list of tablespace names to specify which will be backed up. The tablespace name is translated into a list of data filenames, whether the tablespace has one data file or many. datafile You use the datafile keyword to list specific data files that you want backed up. The data files can be specified by their filenames or by their file number as stored in the database control file. Recovery Manager will include the control file if you specify file 1 (the first file in the SYSTEM tablespace). datafilecopy The datafilecopy specification is a list of data file copies to include in the backup set. Again, the files can be specified by filename or by file number. archivelog In the archivelog specification, you define a filename pattern that will be used to determine which archived logs to include in the backup set. You can also specify the files to back up by a date/time range.

You can specify three other objects when declaring your backup object list: current controlfile Backup controlfile Backupset Use this to back up the current database control file. Use this to include the backup control file. Use this to specify that you're backing up a backup set. The backup set must be disk resident and is specified by its primary key.

BACKUP Command Operands Each operand can be used multiple times within a BACKUP command within your backup script-as long as each one corresponds to a different backup specification: q tag Use this operand to specify a name for this particular backup. Listing 15.1 uses the tag fullback when specifying the incremental backup (level 0). This operand is optional; if it isn't used, it defaults to a null value. q parms This operand is used to pass platform-specific information to the operating system during a backup operation. It takes the form of a quoted character string and is passed each time a backup piece is created. q format This operand is used by the BACKUP command as well as by the ALLOCATE CHANNEL command. As mentioned earlier, defining the filename format here in the BACKUP command overrides the format you may have defined in the ALLOCATE CHANNEL command. Table 15.1 lists the substitution variables that you can use to help create unique filenames. You can use any or all of these substitution values when creating your output file specs. Creating output filenames You use the format operand to define the backup objects' output filenames. It's similar to the way you define archived log filenames in your database init.ora file. Table 15.1 Substitution variables for the format operand Variable%d %p %s %n %t %u Description The database name is put in the file spec The number of the backup piece within the backup set The number of the backup set The database name (padded) A timestamp An eight-character value composed of the backup set number and the time it was created

http://www.informit.com/content/0789716534/element_015.shtml (11 of 16) [26.05.2000 17:05:20]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8
q

include current control file You use this operand to include the control file in the backup set. During the backup operation, Recovery Manager takes a snapshot of the control file and includes the snapshot in the backup set. filesperset This operand allows you to specify how many data files can be included in a single backup set. You can use this operand to control how large the backup set gets and reduce the likelihood of exceeding the operating system maximum file size. When a backup operation reaches the maximum number of files, the backup set is closed and a new one is opened. channel This optional operand allows you to specify which allocated channel to use when writing the backup set. If you exclude this operand, the channels are assigned dynamically by the software. delete input This operand tells Recovery Manager to delete the input files used during the backup operation after the backup is completed. This operand is valid only when you're backing up data file copies or archived log files; it won't work on your active database data files.

The following code lines show the BACKUP command's syntax: backup incremental level 0 tag fullback filesperset 50 format '/ora03/backup/%d/%d__t%t_s%s_p%p' (database);

The COPY Command

You use the COPYcommand to make an image backup on an individual database file. The actual file can be a database data file, a copy of a data file, a current or backup control file, or an archived log file. You can't use the COPY command to write files to tape; they must be to disk. COPY and BACKUP similarities Because the COPY command does essentially the same thing as the BACKUP command, the use of the COPY command is almost identical to that of the BACKUP command. You don't create backup sets with the COPY command, however. The real value of the COPY command is in its capability to multiplex. You can get a large number of files copied in a very short period of time if you list the files to back up individually and create enough channels. The following sample scripts illustrate the COPY command and the parallelization that it can do. When the COPY command is used as it is in Listing 15.2, Recovery Manager copies the files as five separate operations performed sequentially (lines 7-11). It changes things quite a bit if you use only one COPY command and separate the filenames with commas, as lines 7-11 in Listing 15.3 show. Listing 15.2 COPSAMP1.RCV-Using the COPY command with a single thread 01: run { 02: allocate channel 03: allocate channel 04: allocate channel 05: allocate channel 06: allocate channel 07: copy datafile 10 08: copy datafile 11 09: copy datafile 12 10: copy datafile 13 11: copy datafile 14 12: }

d1 d2 d3 d4 d5 to to to to to

type disk; type disk; type disk; type disk; type disk; '/ora01/backup/JDBASE/jdbase10.dbf'; '/ora02/backup/JDBASE/jdbase11.dbf'; '/ora03/backup/JDBASE/jdbase12.dbf'; '/ora04/backup/JDBASE/jdbase13.dbf'; '/ora05/backup/JDBASE/jdbase14.dbf';

Listing 15.3 COPSAMP2.RCV-Image COPY command with parallelization 01: run { 02: allocate channel d1 type disk; 03: allocate channel d2 type disk;

http://www.informit.com/content/0789716534/element_015.shtml (12 of 16) [26.05.2000 17:05:20]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

04: allocate channel d3 type disk; 05: allocate channel d4 type disk; 06: allocate channel d5 type disk; 07: copy datafile 10 to '/ora01/backup/JDBASE/jdbase10.dbf', 08: datafile 11 to '/ora02/backup/JDBASE/jdbase11.dbf', 09: datafile 12 to '/ora03/backup/JDBASE/jdbase12.dbf', 10: datafile 13 to '/ora04/backup/JDBASE/jdbase13.dbf', 11: datafile 14 to '/ora05/backup/JDBASE/jdbase14.dbf'; 12: } By specifying the COPY command as a single operation (Listing 15.3), Recovery Manager will use the five disk channels you allocated all at the same time and back up the five files concurrently. There's a great performance gain in this particular case because the output files are all being written to different disks, and will complete far faster than in the example shown in Listing 15.2.

Restores
There's no point in performing backups if there's no way to return the information to the system in case of a system failure. Recovery Manager provides the RESTORE command to restore the backup sets you created with the BACKUP command, and it's just as easy to use as the BACKUP command. Now it's time to restore the full backup that you performed earlier in the "Recovery Manager Scripting Commands" section.

Restoring the Full Backup

Begin by assuming that some system failure caused corruption or deletion of the database files you backed up earlier. Restoring the files and recovering the database can all be done in a single operation. First, use Server Manager to start your database instance (not mounted, of course): ash$ svrmgrl Oracle Server Manager Release 3.0.4.0.0 - Production (c) Copyright 1997, Oracle Corporation. All Rights Reserved. Oracle8 Enterprise Edition Release 8.0.4.0.0 - Production PL/SQL Release 8.0.4.0.0 - Production

Total System Global Area Fixed Size Variable Size Database Buffers Redo Buffers SVRMGR> exit Server Manager complete.

4749504 47296 4218880 409600 73728

bytes bytes bytes bytes bytes

Starts the instance nomount; it's mounted by the restore script With the database started as NOMOUNT, all you have to do is tell Recovery Manager to restore the entire database. Listing 15.4 shows the Recovery Manager script we will use to perform the restore. (I didn't create a stored script because I wanted you to see how to execute an rman script stored on disk.) Listing 15.4 FULLREST.RCV-Full database restore script using Recovery Manager 01: 02: 03: 04: 05: 06: # fullrest.rcv # This recovery manager script will restore the entire # database as backed up # with the fullback script # run {

http://www.informit.com/content/0789716534/element_015.shtml (13 of 16) [26.05.2000 17:05:20]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

07: 08: 09: 10: 11: 12: 13: }

allocate channel d1 type disk; restore database; sql "alter database mount"; recover database; sql "alter database open"; release channel d1;

That's all it will take to complete the restore and recovery of the database. Notice that you don't even have to specify a location where the backup can be found-this is all handled by Recovery Manager. The following code is a little long but shows all the steps that Recovery Manager took to restore the database and even open it for the users: ash$ rman target jduer/baseball rcvcat recman/recman@rcover cmdfile fullrest.rcv Recovery Manager: Release 8.0.4.0.0 - Production RMAN-06006: connected to target database: jdbase (not mounted) RMAN-06008: connected to recovery catalog database RMAN> # fullrest.rcv 2> # This recovery manager script will restore the entire database as backed up 3> # with the fullback script 4> # 5> run { 6> allocate channel d1 type disk; 7> restore database; 8> sql "alter database mount"; 9> recover database; 10> sql "alter database open"; 11> release channel d1; 12> } 13> RMAN-03022: compiling command: allocate RMAN-03023: executing command: allocate

RMAN-08019: channel d1: restoring datafile 1 RMAN-08509: destination for restore of datafile 1: /ora01/oradata/jdbase/system01.dbf RMAN-08019: channel d1: restoring datafile 2 RMAN-08509: destination for restore of datafile 2: /ora01/oradata/jdbase/rbs01.dbf RMAN-08019: channel d1: restoring datafile 3 RMAN-08509: destination for restore of datafile 3: /ora01/oradata/jdbase/temp01.dbf RMAN-08019: channel d1: restoring datafile 4 RMAN-08509: destination for restore of datafile 4: /ora01/oradata/jdbase/tools01.dbf RMAN-08019: channel d1: restoring datafile 5 RMAN-08509: destination for restore of datafile 5: /ora01/oradata/jdbase/users01.dbf RMAN-08023: channel d1: restored backup piece 1 RMAN-08511: piece handle= /ora03/backup/JDBASE/JDBASE__t331940454_s3_p1 params=NULL RMAN-08024: channel d1: restore complete

http://www.informit.com/content/0789716534/element_015.shtml (14 of 16) [26.05.2000 17:05:20]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

RMAN-03022: compiling command: recover RMAN-03022: compiling command: recover(1) RMAN-03022: compiling command: recover(2)

RMAN-03022: compiling command: recover(4) RMAN-03022: compiling command: sql RMAN-06162: sql statement: alter database open RMAN-03023: executing command: sql RMAN-03022: compiling command: release RMAN-03023: executing command: release RMAN-08031: released channel: d1 Recovery Manager complete. Disk channel d1 is created and allocated RESTORE DATABASE command is executing Newly restored database Database data files are recovered, completing the restoration

More on Restores
The restore script in Listing 15.4 uses the sql command to mount the database and then to later open it. This very powerful Recovery Manager feature allows for incredible flexibility and programmability of many backup and restore scenarios. For example, if you wanted to recover a tablespace while the database was open, the meat of your restore script would be only three lines: Sql "alter tablespace JDDATA offline"; Recover tablespace JDDATA; Sql "alter tablespace JDDATA online"; The rest of the database would remain available while the restoration was proceeding, and the tablespace in question would be made available again before Recovery Manager exited. Another note on restores is Recovery Manager's capability to perform point-in-time recovery. This is done by using the SET UNTIL TIME command in your restore script. Suppose that today was May 1, 1998, and there was a database problem at 2:01 p.m. that required recovery of one of the database objects. Your restore script would essentially be the same, except for the addition of this command: Set until time '1-MAY-1998 14:00:00' Point-in-time recovery Oracle documentation dedicates an entire chapter to point-in-time recovery, and there are many pre-requisites to performing this operation. Consult the documentation and call Oracle Support for more information; a full discussion is beyond the scope of this book. Executing this script would restore the database object to the condition it was in at 2:00 p.m. that day.

Using the REPORT and LIST Commands

You use the REPORT command to dump information out of the recovery catalog. This command can tell you which database objects need to be backed up, which backup sets are obsolete, and so on.

http://www.informit.com/content/0789716534/element_015.shtml (15 of 16) [26.05.2000 17:05:20]

informit.com -- Your Brain is Hungry. InformIT - Using Recovery Manager for Backup and Recovery From: Using Oracle8

The LISTcommand is used to report on the status of backup sets and the like. For example, you use the rman command LIST BACKUPSET OF DATABASE to list information on the full backup script and when it was performed: RMAN> list backupset of database; RMAN-03022: compiling command: list RMAN-06230: List of Datafile Backups Key File Type LV Completion_Ckp ----- ---- ------------ -- -------------27 1 Incremental 0 29-APR-98 27 2 Incremental 0 29-APR-98 27 3 Incremental 0 29-APR-98 27 4 Incremental 0 29-APR-98 27 5 Incremental 0 29-APR-98

SCN ----17900 17900 17900 17900 17900

Ckp Time --------29-APR-98 29-APR-98 29-APR-98 29-APR-98 29-APR-98

The REPORT and LIST commands' full syntax is spelled out on the Oracle documentation CD-ROM. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_015.shtml (16 of 16) [26.05.2000 17:05:20]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Using Optimizers and the Analytic and Diagnostic Tools

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

Functions of the Optimizer

r r r r r r r

Simplifying and Transforming Statements Choosing a Rule-Based or Cost-Based Optimization Approach Data Access Paths Table Join Options Using Rule-Based Optimization Using Cost-Based Optimization Using Hints to Influence Execution Plans Creating a Plan Table Using the EXPLAIN PLAN Command Interpreting the EXPLAIN PLAN Results Creating a Trace File Formatting a Trace File with TKPROF Interpreting Trace Information Controlling the EXPLAIN Option Output The Statistics Option Output A Sample Session with AUTOTRACE

The EXPLAIN PLAN Utility

r r r

The Trace Facility

r r r

Using AUTOTRACE
r r r

Functions of the Optimizer

The Oracle optimizer is responsible for identifying the best means to execute any SQL statement that needs to be processed. This includes a number of tasks: Why the Optimizer simplifies expressions and statements You can write a SQL statement to perform a specific task in many ways, the variety increasing as the numbers of conditions and different tables increase. However, you have a very limited number of choices of retrieval paths to a table's rows and to methods for joining multiple tables. The number of possible matches that must be checked is reduced if statements can be converted to a common set of structures.
q q

Simplifying expressions and transforming statements where possible Choosing a rule-based or cost-based optimization approach

http://www.informit.com/content/0789716534/element_016.shtml (1 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8
q q

Choosing an access path to retrieve the data from each table Developing a join strategy, which includes defining the order in which the tables will be joined and the join operation to be used for each join.

Simplifying and Transforming Statements

The optimizer automatically simplifies certain common constructs in SQL statements if the result will simplify the execution of the statement. Such conversions can range from the very simple, such as simplifying the expression 2000/10 to the integer 200, to transforming a statement with an OR operation into a compound query with two component queries. The former type of simplification will always be done; the latter will depend on whether there are indexes on the columns in the original WHERE clause and which optimizer approach is being used. Many different types of transformations can occur, including, but not limited to, the following: q Simplifying arithmetic expressions q Converting the IN operator to a series of OR conditions q Converting a BETWEEN...AND clause to a pair of comparison expressions q Integrating a view's definition into the statement's conditions q Transforming an OR operation into a compound query q Transforming a complex statement into a single statement with a join condition The Oracle8 Server Concepts Manual contains detailed explanations of these and all other possible statement simplifications and transformations. You should be aware of them mainly to understand why the optimizer may sometimes choose an execution plan that doesn't appear to be appropriate for the structure of the statement being processed.

Choosing a Rule-Based or Cost-Based Optimization Approach

Oracle chooses an optimization approach for each statement based on several criteria: 1. Referenced objects having a defined degree of parallelism 2. Hints in the statement 3. Session setting of OPTIMIZER_GOAL 4. Value of initialization parameter OPTIMIZER_MODE 5. Existence of statistics on referenced objects SEE ALSO Find out more about database parameters and the initialization file, Figure 16.1 shows how the optimizer uses these criteria to determine whether it uses rule-based or cost-based optimization. The first factor in choosing the optimization approach, when examining a statement to be processed, is to see if it can be executed in parallel. This includes determining if there will be at least one full table scan and if any of the objects referenced by the statement were defined with the PARALLEL option. If both conditions are true, Oracle will use cost-based optimization to create an execution plan containing parallel steps. Figure 16.1 : Oracle examines many factors to determine whether the optimization approach should be rule-based (RBO) or cost-based (CBO). Optimization for parallel processing and statements with hints Oracle must use cost-based optimization to develop execution plans for statements that require parallel server processes, because the rule-based optimization approach has never been upgraded to handle parallel execution. Similarly, the only optimization option that can interpret hints is cost-based. Unless the only hint is RULE, Oracle uses cost-based optimization for any statement with one or more hints. For statements that can't be executed, at least partially, in parallel, Oracle looks to see whether the statement contains a hint. If any hint exists (other than the RULE hint), cost-based optimization will be used to process the statement. For statements that don't require cost-based optimization because of parallel execution steps or the presence of hints, Oracle checks to see if the session has defined an optimization choice for itself. The command ALTER SESSION SET OPTIMIZER_GOAL = goal

http://www.informit.com/content/0789716534/element_016.shtml (2 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

Where you can use the RULE keyword RULE is the option name (as well as the name of the optimizer approach) used in three different types of syntax to denote the type of optimization required. It's used in hints, in the ALTER SESSION command, and in the database initialization file. It's the same RULE in each case, but in the first situation, it is known as a RULE hint; in the second, it's a RULE goal (because the command sets the OPTIMIZER_GOAL parameter), and in the last case, it's the RULE mode (the parameter name is OPTIMIZER_MODE). allows you to choose from one of the four optimization goals: FIRST_ROWS, ALL_ROWS, RULE, and CHOOSE. The first two options cause Oracle to use cost-based optimization, FIRST_ROWS causing it to find the execution plan that will return the first row as quickly as possible and ALL_ROWS optimizing the overall response time of a statement. The RULE goal always causes rule-based optimization to be used for the session by default. The final option, CHOOSE, as you can see from Figure 16.1, can cause rule-based or cost-optimization to be selected, based on the existence of statistics. The statistics used to decide which optimization approach will be selected are collected with the ANALYZE command (discussed in the "Collecting Statistics for the Cost-Based Optimizer" section later in this chapter). When Oracle needs to optimize a statement running in a session with its optimizer goal set to CHOOSE, it looks in the data dictionary to see if any one of the segments referenced in the statement have statistics. If statistics are found, the statement will be processed with cost-based optimization, which uses these statistics to help develop its execution plan. If there are no statistics on all the segments to be processed, rule-based optimization is used. When Oracle has no other indications as to which optimization approach to take, it uses the value assigned to the initialization parameter, OPTIMIZER_MODE. The same four values-FIRST_ROWS, ALL_ROWS, RULE, and CHOOSE-can be assigned to this parameter; they act in the same way as discussed for the ALTER SESSION command. By default, the parameter is set to CHOOSE, which means that the optimization approach chosen for any given statement will depend on the status of the statistics in the data dictionary for the segments processed by the statement. It also means that the optimization approach can change if statistics are added or dropped over time. Choosing FIRST_ROWS versus ALL_ROWS The FIRST_ROWS option is best used for statements executed in interactive applications, because users of such applications are typically waiting for responses from the system as soon as they initiate a process. Even if the overall execution time isn't minimized, a user can probably begin doing useful work with the first row of data returned, so the delay while the remainder of the data is processed isn't detrimental. ALL_ROWS should be used when the statement needs to execute as quickly as possible. You should always choose this option when initializing the optimizer for a batch program or for a program that may otherwise have unacceptably poor response time.

Data Access Paths

Table 16.1 lists the access paths available to reach the required rows in a table. The Rank column is included for a later discussion about rule-based optimization. Table 16.1 Optional data access paths to be evaluated by optimizer Access Path: Bitmap Index Scan Fast Full Index Scan Single Row by ROWID Single Row Cluster Join Single Row Hash Cluster Description: Accesses via a bitmap index entry Performs a full scan on the index entries rather than on the table Uses the rowid as provided by a current cursor or a WHERE clause with a rowid value Returns only a single row from two or more tables in a cluster with a join condition on the cluster key Returns a single row from a hash cluster when the WHERE clause identifies the complete hash key, which is also a unique or primary key Rank: Not ranked Not ranked 1 2 3

Single Row by Key

Clustered Join

Returns a single row from a table when the WHERE 4 clause identifies all columns in a unique or primary key Returns one or more rows from two or more tables 5 in a cluster with a join condition on the cluster key

http://www.informit.com/content/0789716534/element_016.shtml (3 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

Hash Cluster Key Index Cluster Key Composite Index Single-Column Index(es) Bounded Range Index Search

Returns one or more rows via the cluster-key value 6 Returns one or more rows via the cluster-key value 7 Returns one or more rows when all columns of a composite index are referenced Uses one or more single-column indexes 8 9 10

Uses a single-column index, or the leading column(s) of a composite index, to find values in a bounded range (with a lower and an upper value) Unbounded Range Index Search Uses a single-column index, or the leading column(s) of a composite index, to find values in an unbounded range (with a lower or an upper value, but not both) Sort-Merge Join Join of two tables via a join column when the tables aren't clustered together MAX or MIN of Indexed Column Returns the column maximum or minimum value from an index if the column is indexed by itself or is the leading column of a composite index, if the query has no WHERE clause, and if no other column is named in the SELECT clause ORDER BY on Indexed Column Uses single column index or leading column of a composite index to find rowids of table rows in order when column is guaranteed not to contain NULLs Full Table Scan Reads rows directly from a table

11

12 13

14

15

Rank column used in rule-based optimization only The two access paths that show the value "Not ranked" can be used only by the cost-based optimization approach; therefore, they have no rank value for rule-based optimization. Using an index to find column minimum or maximum values An index provides a convenient access path for a maximum or minimum value of a column because the entries are sorted from least (the first entry is the minimum) to greatest (the last entry is the maximum). If the query needs other columns or has other restrictions on which rows are required, this retrieval path is inappropriate because it can't identify any other rows that must be considered. Also, the index can be used only if it's the leading column of a composite index or is the only column in the index.

Table Join Options

A table join occurs when the rows from a table are combined with the rows from another table, or even with the rows from itself. The latter, known as a self-join, is used when the value in one column needs to be matched with the same value in another column. Joins typically involve matching a value in a column, or set of columns, in one table with a corresponding value, or set of values, in the other table. When a value in each table must match, the resulting join is known as an equijoin. If the join condition is based on an inequality between the columns, the join is called a non-equijoin. Other join options are Cartesian products, which occur when there's no controlling condition and every row in one table is joined to every row in the other table, and outer joins, which include all rows from one table even if there's no matching value on the join column(s). Consider a table in a manufacturing application. The application records information about individually manufactured parts as well as assemblies-that is, combinations of parts. For example, this book you're reading is an assembly of pages (one type of part) and a set of covers (another type of part). The series, of which this book is a part, is an assembly of the individual books that comprise it. A database table containing assembly information might include, among others, the columns ASSEMBLY_ID and PART_ID, where PART_ID contains a part identification value and ASSEMBLY_ID is the code number for an assembly that contains one or more parts. If we selected the rows for assembly ID 10-1-AA, we might see the following: PART_NUMBER 10-1-AA 10-1-AA 10-1-AA PART_NAME 12-8HY-U-87 9JD7-RT-9 LK-LG-55624

http://www.informit.com/content/0789716534/element_016.shtml (4 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

Left and right joins Outer joins, when a table having no matching data has its rows included in the result set anyway, are some-times referred to as left and right joins. Depending on whether the join condition lists the column of the non-matched table on the left or right side of the WHERE condition, the join is considered a left or a right join. Although Oracle doesn't allow left and right joins in a single statement, it will allow a view based on a left join to be included in query with a right join, and vice versa. If we really wanted to see the part names as well as the part numbers for the parts that comprise assembly 10-1-AA, we would need to code a self-join: SELECT a.part_number, p.part_number, p.part_name FROM assemblies a, assemblies b WHERE a.part_number = '10-1-AA' AND a.part_name = p.part_number; Oracle performs joins in a number of different ways, as summarized in Table 16.2. Table 16.2 Oracle chooses a join method from among a number of options Join Operation: Nested Loops Sort-Merge Cluster Join Hash Join/1/ Star Query/1/, /2/ Star Transformation/1/,/2/
/1/ /2/

Characteristics: For each row retrieved from the driving table, looks for rows in the driven table Sorts the rows from both tables in order of the join column values and merges the resulting sorted sets For each row retrieved from the driving table, looks for matching rows in the driven table on the same block Builds a hash table from the rows in the driving table and uses the same hash formula on each row of the driven table to find matches Creates a Cartesian product of the dimension tables and merges the result set with the fact table Uses bitmap indexes on the dimension tables to build a bitmap index access to the fact table

Method available only when using cost-based optimization. Any of the other options can be used to join the dimension tables and join that result to the fact table.

When two tables need to be joined, the optimizer evaluates the methods as well as the order in which the tables should be joined. The table accessed first is the driving table; the one accessed next is the driven table. For joins involving multiple tables, there's a primary driving table, and the remaining tables are driven by the results obtained from the previous join results. Two situations will always cause the optimizer to select a specific table order when performing table joins: q If a table is guaranteed to return just one row based on the existence of a unique or primary key, this table will be made the driving table. q If two tables are joined with an outer join condition, the table with the outer join operator-a plus sign enclosed in parentheses, (+)-will always be made the driven table of the pair.

Using Rule-Based Optimization

Although Oracle8 still supports the rule-based optimizer, Oracle Corporation strongly encourages you to migrate to cost-based optimization. The rule-based optimizer won't be included in later releases of the database, although it isn't clear just when this will be. Support is being maintained to allow customers time to complete the transfer, tuning, and implementation of their applications-in-house and third party-to cost-based optimization. Features not available to the rule-based optimizer

http://www.informit.com/content/0789716534/element_016.shtml (5 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

Some features introduced in recent releases can't be used by the rule-based optimizer. These features include partitioned tables, index-only tables, reverse-key indexes, bitmap indexes, parallel queries, hash joins, star joins, star transformations, histograms, and fast full index scans. When the feature is an aspect of an object, such as a reverse-key index, the rule-based optimizer will act as though the object weren't avail-able for use. In cases where there's no choice but to use the feature (such as a index-only table), Oracle will automatically use the cost-based optimizer. Other optional features-such as having a default degree of parallelism on a table-will cause Oracle to use the cost-based optimizer to take advantage of the feature. If you're still using rule-based optimization, you should be planning your strategy to convert to the cost-based approach. You can't take advantage of many new features of the database while using the rule-based optimizer, and you may find that you can improve performance by using the newer optimizer without having to implement these new features, should you not have the resources to investigate them. Beginning, or at least anticipating, the conversion now, before there's a concrete deadline you have to meet, should help you realize a better product overall. If you haven't set the value of the OPTIMIZER_MODE parameter in your initialization file and haven't executed the ANALYZE command to collect statistics for any of the tables, indexes, or clusters used in applications, your applications are probably running against the rule-based optimizer. However, you can't guarantee this because q Application developers and end users could include hints in the SQL statements. q Applications and user sessions could set the OPTIMIZER_GOAL. q Segments could be defined with a default degree of parallelism. The rule-based optimization approach is so named because it follows a standard set of tests when determining what access path to use to obtain the rows required for each step of a statement's execution. Table 16.1 earlier in this chapter shows the possible access paths to a table and includes a rank number to show which approaches are preferred. During rule-based optimization, the table is tested to see if it can be accessed by each access path in turn, beginning with the rank 1 option, and the first possible path is chosen as the access path. A good reason to continue using rule-based optimization Although cost-based optimization is becoming the preferred approach, you shouldn't abandon the rule-based approach, if that is what you have been using, without due consideration. Poor performance can result if your database is using cost-based optimization without any statistical information from which to derive good execution plans. Statistics that are no longer current can also be a detriment to cost-based optimization. If the statement requires a table join, the rule-based approach uses an algorithm to determine the two key elements of the join: first, which will be the driving table and which the driven table; and second, which join method will be used. The rules of this algorithm are as follows: q Choose each table in turn as the driving table and build a possible execution plan for each one. q In each potential execution plan, add the other tables in order by rank, the lower the rank number, the closer to the driving table. q Choose a join method for each driven table by looking at its rank number: r If its rank is 11 or better, use nested loops r If its rank is 12 or lower and r There's an equijoin condition in join order (such as an index on the join column(s)), use sort-merge r There's no equijoin, or equijoin in join order, use nested loops q Select the resulting execution plan with the least nested-loop operations with the driven table being accessed via a full table scan. q If there's a tie between two or more execution plans, select the plan with the least sort-merge operations. q If there's still a tie, select the plan with the best (lowest numbered) ranked access path for its driving table. q If there's still a tie, choose the plan with the most merged indexes for access to the driving table, or else the one that uses more of the leading columns of a concatenated index. q If this still results in tie, use the plan that uses the last table named in the FROM clause as its driving table. Changing the execution plan under rule-based optimization

http://www.informit.com/content/0789716534/element_016.shtml (6 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

If you don't think the rules will generate the best execution plan for a given statement under rule-based optimization, you can try to improve it. For instance, to stop the optimizer by using an index, you can modify the reference to the column in the WHERE clause by adding a NULL or zero-length string for character columns (such as USERNAME || ''), or a zero for numeric columns (such as CUSTOMER_ ID + 0). This won't change the results returned but will prevent use of the index. To force the use of an index that's being ignored, you may have to rewrite the statement to avoid modifying the column reference, such as removing functions such as UPPER.

Using Cost-Based Optimization

As mentioned in the previous section, the cost-based approach will eventually be the only optimization approach available. Meanwhile, it will continue to support new database features that the rule-based approach can't handle. You should plan to convert your applications-if you haven't already-to run under cost-based optimization. In Figure 16.1, you can see that you can invoke the cost-based optimization in a number of ways. You can do it directly with the FIRST_ROWS or ALL_ROWS setting in the initialization file or in an ALTER SESSION command; you can do it less directly by defining a default degree of parallelism on a segment being accessed in a statement, or by including a hint (other than RULE) in a statement. If you allow Oracle to use its default behavior to select an optimization approach (which means using the CHOOSE option for the OPTIMIZER_MODE parameter), the choice will be based on the existence or non-existence of statistics on the segments referenced in the statement. The reason for the optimizer approach to depend on statistics is quite reasonable. Cost-based optimization depends on these statistics to compute the relative costs of performing different execution plans in order to choose the most efficient. If the statistics aren't stored, the likelihood that the optimizer will choose a good plan is reduced significantly, and Oracle would prefer to rely on the rule-based approach. Of course, if you force the optimizer approach to be cost-based (with a hint, for example), it will work-albeit poorly-in the absence of statistics. It may also perform less than optimally if the statistics it uses are stale and no longer reflect the true nature of the segment they're meant to describe. It's therefore essential that you be prepared to maintain current statistics on the database segments if you want to use cost-based optimization. If you fail to do that, your application developers-and even end users who access the database directly-will have to include hints in many of their statements to ensure that a reasonable execution plan is used. Statistics are generated and maintained with the ANALYZE command. The syntax for the options related to cost-based statistics is as follows; the other options are covered in other chapters where they are relevant. ANALYZE TABLE|INDEX|CLUSTER [schema.]table_name|index_name|cluster_name [PARTITION (partition_name)] COMPUTE|ESTIMATE|DELETE STATISTICS [table_clause][,...] [SAMPLE integer [ROWS|PERCENT] What you should analyze The COMPUTE option of the ANALYZE command generally consumes more resources than the ESTIMATE option and, consequently, can have a greater impact on your application performance. You should compare the results from both options, using different estimated sample sizes, to decide if you really need to incur the extra over-head of computing exact statistics. When analyzing tables with associated indexes, you also can reduce the work performed by the database by estimating the table statistics and then computing exact values on the indexes individually. This will provide the optimizer with the best statistics when accessing the data via the indexes, which is how the data should be retrieved if the indexes are doing their job. where
q q q

ANALYZE TABLE, INDEX, and CLUSTER identify the type of segment to be analyzed. schema. is required if the segment belongs to another user. table_name is the name of the table being processed. It's required if the keyword TABLE is included in the command. index_name is the name of the index being processed. It's required if the keyword INDEX is included in the command. cluster_name is the name of the cluster being processed. It's required if the keyword CLUSTER is

http://www.informit.com/content/0789716534/element_016.shtml (7 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

included in the command. PARTITION (partition_name) identifies the partition name if a single table or index partition is being analyzed. This option isn't valid when analyzing clusters. COMPUTE|ESTIMATE|DELETE STATISTICS identifies which operation you want to use. You must include one of the three operations when the keyword STATISTICS is included in the command. COMPUTE provides exact statistics, ESTIMATE uses a sample of the data to generate statistics, and DELETE removes any previously analyzed statistics. SAMPLE integer sets the size of the sample used in the estimate. It can be used only with the ESTIMATE option. The sample size will default to 1,064 rows if the ESTIMATE clause is used without the SAMPLE option. ROWS|PERCENT indicates if the sample value should be treated as a row count or a percentage of the table size. It can be used only with the ESTIMATE option. table_clause is allowed only when the segment being analyzed is a table and you are using the COMPUTE or ESTIMATE option. The format of the table_clause is

Specifies that the command will create table statistics only; no column or index statistics will be generated Specifies that the command will create histogram statistics on every column Specifies maximum number of buckets in the histogram; default value is 75 if option isn't included Specifies that the command will create histogram statistics only on indexed columns Specifies that the command will create histogram statistics on the named column(s) or object scalar type(s) Specifies that the command will create statistics on every indexed column, but not on the table Specifies that the command will create statistics on every local index partition; must be included if FOR ALL INDEXES and PARTITION options are specified The TABLE options that create histograms should be used if your table has a very uneven distribution of values in columns used for retrieval. When different values are stored in a column, the optimizer assumes that they will each appear about the same number of times. If some of the values occur only rarely and one or two of the others occur in a large proportion of the records, this assumption may not lead to a good execution plan. The frequently occurring values should be accessed by a full table scan, whereas the infrequently appearing values would be best retrieved via an index. By building a histogram, you provide the optimizer with the information it needs to distinguish between these two types of values and assist it in building a good execution plan. The number of buckets, or partitions, in the histogram determines how finely the different values are distinguished. The more buckets, the greater the chance that the histogram will show the frequency of occurrence of any specific value in the column. If you need to isolate only one or two disproportionately occurring values, however, you need fewer buckets. You can use the ANALYZE command to recalculate statistics any time you want without having to delete the old ones first. You should plan to perform re-analysis on a regular basis if the segment changes frequently. Keeping statistics current

http://www.informit.com/content/0789716534/element_016.shtml (8 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

You should monitor the statistics on your database segments to make sure that they stay current. I recommend that you begin by executing the ANALYZE command and recording the statistics from the related view: DBA_TABLES, DBA_INDEXES, or DBA_ CLUSTERS. Re-execute the ANALYZE command a month later and compare the new statistical values; if they're close in value to the previous month's, you shouldn't need to perform another analysis for a few more months. If the statistics are very different, you may need to check again in a week. If they're somewhat different, you should plan to re-analyze the table every month. Over time, you should develop a sense of how frequently each different segment needs to be analyzed. You may need to run a program once a week, once a month, or at some other fixed interval. Your program may analyze just a few segments each time it's run, with additional segments every other time, more every third or fourth time, and so on. When a statement is processed with cost-based optimization, the execution plan will include the table selection access paths and join methods based on the lowest estimated costs. These costs take into account the number of Oracle blocks that have to be manipulated, the number of reads that may need to occur to retrieve these blocks from disk into memory, the amount of additional memory that may be needed to process the data (such as space to complete sorts or hash joins), and the cost of moving data across any networks. If you've built your database objects with application schemas-that is, where all the objects belonging to an application are owned by the same user-you can simplify the task of collecting statistics for cost-based optimization. Oracle provides a procedure, ANALYZE_SCHEMA, in its DBMS_UTILITY package, which will run the ANALYZE command for you against every segment in a named schema. If you haven't already done so, you need to execute the CATPROC.SQL script, which you can find in the admin subdirectory of your ORACLE_HOME directory, as SYS to build the necessary PL/SQL structures. You can then execute the required procedure by using SQL*Plus's EXECUTE command or by creating your own PL/SQL routine to run the procedure. The SQL*Plus EXECUTE command would look like this: View Code SEE ALSO Information about the various Oracle-supplied SQL scripts mentioned in this chapter, You would substitute the name of the schema holding the segments you want to analyze at the username prompt; the COMPUTE, ESTIMATE, or DELETE keyword at the option prompt; and a number, the keyword NULL, or an empty string ('') for the rows and pct prompts. The last two options are relevant only for the ESTIMATE option, and any values provided are ignored for other options. They indicate the number of rows or the proportion of the table to be included in the sample respectively. If you don't provide a number for either, or set both to zero, the sample uses the default number of rows (1,064). If you provide a number for both, the value for rows is used unless it's zero, in which case the percentage sample size is used. The statistics collected with the ANALYZE command are used in computing these costs. In cases where the cost-based optimizer is being used for a statement that references one or more-or even all-segments that have no statistics available, it still has to evaluate the potential costs of different execution plans. To do this, it uses basic information from the data dictionary and estimates the missing values. Naturally, the results aren't as accurate as they would be with current statistics collected with the ANALYZE command.

Using Hints to Influence Execution Plans

To overcome poor execution plans-due to missing or out-of-date statistics, or even due to unusual distribution of data in a table or index not anticipated by the optimizer-you can include hints in a statement. Hints are similar to the "tweaks" I suggested you can use to try to modify the behavior of rule-based execution plans, but they're more sophisticated and give you a much wider range of options. Table names in hints If you use a table alias in the FROM clause of a statement, you must also use that alias in the hint string when referencing the table. Your statement won't fail if you fail to do this, but the hint will be treated as comment text and won't be acted on as you expected. Oracle publishes a complete list of available hints with descriptions of what they do and how to use them in the Oracle8 Server Tuning manual, so I won't reproduce that data here. I do include the details required to include a hint in a statement, as this can be confusing: q Place the hint immediately after the SELECT, UPDATE, or DELETE command statement keyword. q Open the hint with a comment delimiter and a plus sign concatenated to it. You can use either form of comment delimiter supported by Oracle: r A forward slash and asterisk: /*+
http://www.informit.com/content/0789716534/element_016.shtml (9 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8
r r

A pair of hyphens: --+ Include a valid hint or series of hints, with no punctuation (other than the required spaces) between adjacent hints. Optionally include comments. Invalid hints and conflicting hints are treated as comment text and ignored. You won't receive an error message for an invalid hint, but the statement will proceed ignoring the intended hint. Terminate the hint comment with a string consisting of s An asterisk and a forward slash-*/- if the hint comment was opened with a /*+ s A carriage return if the hint comment was opened with a double hyphen: --+ A hint enclosed with /*+ ... */ can span multiple lines, whereas a hint introduced with --+ is always terminated at the end of a line.

Understand the difference between hint delimiters

Here is an example of a hint using some of the features just discussed:

Use comment delimiter to begin hint Add the plus sign to indicate this string will contain hints Use the hint name in the string Separate additional hints with at least one space

The EXPLAIN PLAN Utility

To observe the execution plan that a particular statement would use were it to be executed, you can use Oracle's EXPLAIN PLAN feature. With this utility, you can examine the execution plan of statements you think may run inefficiently and determine if they're using appropriate access paths; you can observe the impact of changing a statement to use, or to avoid using, specific indexes; you can check the execution plans used when including different hints in the same statement; and you can even see if re-analyzing a segment to gather newer statistics or a different sample size results in a different execution plan. To use the EXPLAIN PLAN feature successfully, you'll need to r Create a special table to hold the execution plan. r Know how to use the EXPLAIN PLAN command. r Learn how to interpret the results.

Creating a Plan Table

The easiest way to create a table to hold the results of an EXPLAIN PLAN command is to run the Oracle script UTLXPLAN.SQL. You can find this script, along with the others mentioned in this chapter, in the admin subdirectory under your ORACLE_ HOME directory. This will build a table, named PLAN_TABLE, in the schema of whichever user is executing the script. Create a different PLAN TABLE for different users Consider creating this table under every userid that might be used to evaluate statement performance characteristics by executing UTLXPLAN.SQL for each one. This will allow users to use the EXPLAIN command without having to include the table's schema in the command, or use a synonym to identify a common table. It also will automatically provide a table with the required privileges for use with the EXPLAIN command. Finally, it will allow users to manage the contents of the table for themselves, reducing the possibility of confusing different execution plans in a shared table. If you want, you can create just one plan table and grant the required privileges on it to any other users who

http://www.informit.com/content/0789716534/element_016.shtml (10 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

may need to execute the EXPLAIN PLAN command. Minimally, you should grant the INSERT, SELECT, and DELETE privileges if you want the table to be shared. Although not recommended, you can create your own plan table by hand or change the name of the table to something other than PLAN_TABLE. If you do the latter, you have to include the name in a number of commands that would otherwise use the default name, and you can't use all the features of the AUTOTRACE feature (discussed later). If you build the table by hand, you must ensure that you include the identical column definitions from the ULTXPLAN.SQL script.

Using the EXPLAIN PLAN Command

To see the execution plan for a SQL statement, you use the EXPLAIN PLAN command, which has the following syntax: EXPLAIN PLAN [SET STATEMENT_ID = 'label'] [INTO [schema.]table_name[@dblink] FOR statement where
r r

EXPLAIN PLAN and FOR are the required keywords. SET STATEMENT_ID are optional keywords, required only if you want to flag every row in the plan table with an identifier. label is an arbitrary string, up to 30 characters long, that you can use to "label" every row in the plan table generated by the current command. INTO is the option you need to include if the plan table you're using to store your results isn't in your own schema, isn't named PLAN_TABLE, or isn't in your local database. schema is the name of the owner of the plan table you want to use. Your own schema will be targeted if you don't include this option. table_name is the name of the plan table you want to use. You must include this name if you use the INTO option. @dblink optionally connects you to a remote database schema, based on the information in the database link, dblink, for you to use a plan table at that location. By default, your local database will be used. statement is any valid SELECT, INSERT, UPDATE, or DELETE statement for which you want to examine the execution plan. If you're sharing a plan table with other users or want to keep the execution plans for a number of different statements (or versions of the same statement), you need to be able to identify which row belongs to which execution plan. As a regular relational table, the plan table won't necessarily store related rows together but will intermix the rows from different execution plans. Use the statement identifier clause, SET STATEMENT_ID, to include a unique string for each statement you explain, which you can then use to identify the rows associated with its execution plan.

Using a plan table identifier

The statement you examine in an EXPLAIN PLAN command is never executed. It's therefore possible to run the command against an empty table and still see the potential execution plan, although cost-based optimization results may be misleading due to the lack of statistics reflecting real contents. You can also safely rerun the EXPLAIN PLAN command multiple times for a statement that would generate extensive overhead if it were actually to run. This is particularly useful when trying to tune a query against massive data warehouse tables (such queries can run for hours, even days) before submitting it for execution. Privileges needed for the EXPLAIN PLAN command Although the statement identified in the EXPLAIN PLAN command won't be executed, you must have the necessary privileges to run the statement for its execution plan to be generated. You must also have the privilege to INSERT into the plan table. If you don't, you'll receive the same error message as if you tried to execute the statement directly.

Interpreting the EXPLAIN PLAN Results

Table 16.3 shows the meanings of each operation and options that can appear in the plan table after you execute an EXPLAIN PLAN command.
http://www.informit.com/content/0789716534/element_016.shtml (11 of 23) [26.05.2000 17:07:05]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

TABLE 16.3 Operations and options generated by an execution plan Operation: AND-EQUAL Option: Description: An operation that accepts multiple sets of rowids from single-column indexes on the same table and returns the rowids common to all the sets. CONVERSION TO ROWIDS converts the bitmap representation to actual rowids in the table. FROM ROWIDS converts rowids into a bitmap. COUNT returns the number of rowids represented by the bitmap. SINGLE VALUE looks for a single value in the bitmap; RANGE SCAN looks for a range of values; FULL SCAN examines the entire bitmap index. MERGE Merges two or more bitmaps into a single bitmap. Subtracts the bits of the bitmap for a negated predicate from another bitmap. Computes the Boolean OR of two bitmaps. Orders rows for a query containing a CONNECT BY clause. An operation that returns all the rows from two or more sets of rows. An operation to count the number of rows retrieved from a table. STOPKEY FILTER FIRST ROW FOR UPDATE INDEX A count operation that's terminated by a ROWNUM expression. An operation that removes a subset of the rows from a set.

BITMAP

INDEX

MINUS OR CONNECT BY CONCATENATION COUNT

Retrieves the first row only from a query. An operation that locks the rows retrieved when the query contains a FOR UPDATE clause. UNIQUE SCAN An index retrieval guaranteed to find no more than one entry. An index retrieval on a non-unique value or a range of unique or RANGE SCAN non- unique values. A range scan performed in descending DESCENDING order. CONCATENATED Repeats an operation based on the values found in the inlist. RANGE SCAN An operation that combines two sets of rows and eliminates any duplicates. A table join performed by matching values in the tables' join column(s) after they've been sorted. A merge join operation used to perform an outer join. An operation that removes rows from a set of records when they appear in a second set. A table join that compares each value found in one table with values in the second table and returns rows with matching values in the join column(s). A nested loops operation used to perform an outer join. An undocumented internal operation typically involving views. Retrieves data from a remote database. An access of values in a sequence generator. AGGREGATE UNIQUE GROUP BY JOIN ORDER BY A sort performed to apply a group function. A sort performed to eliminate duplicates. A sort performed to satisfy a GROUP BY clause. A sort performed in preparation for a merge-join operation. A sort performed to satisfy an ORDER BY clause. A retrieval that accesses all the rows of a table. A retrieval from a table in an indexed cluster based on a value in the cluster index.

INLIST INTERATOR INTERSECTION MERGE JOIN

OUTER MINUS NESTED LOOPS

OUTER PROJECTION REMOTE SEQUENCE SORT

TABLE ACCESS FULL CLUSTER

http://www.informit.com/content/0789716534/element_016.shtml (12 of 23) [26.05.2000 17:07:06]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

HASH BY ROWID UNION VIEW

A retrieval from a table in a hash cluster based on a hash-key value. A retrieval from a table based on the rowid(s) of one or more rows. An operation that combines two sets of rows and returns all the rows from both sets, other than duplicates. Executes a view's query.

Operations versus options in the plan table The operations named in an execution plan are the actual steps per-formed to process the statement. Options describe why or how the operation is being executed. The results of the operation may or may not be different because of the option. For example, the INDEX operation will always return a set of rowids, possibly empty if no rows match the desired criteria, regard-less of the option used. On the other hand, a SORT operation may only return one row for an AGGREGATE or GROUP BY option, or may return all the rows in the set as when used for the ORDER BY option. Other columns that you may need to review to understand what the execution plan is doing include the following: r OBJECT_NAME The name of the table or index being operated on r OPTIMIZER The current mode or goal of the optimizer r OTHER_TAG Indicates if operations are being performed in parallel with parallel server processes r COST The relative cost as evaluated by cost-based optimization r CARDINALITY The number of rows that the cost-based optimization estimated will be accessed by the operation r BYTES The number of bytes that the cost-based optimization estimated will be accessed by the operation The other useful columns in the plan table are the ID, PARENT_ID, and POSITION columns. Although you can work out the order of operations in the execution plan by using the values in these columns, you'll find it easier to use them to build a tree-walk output from the plan table, using indentation or other means to show the order in which operations will occur and which operations depend on others. For example, if you explain a query that retrieves a single row through a primary key value, you'll have three entries in your plan table. One will show the primary key index access, one will show the table access using the rowid from the index, and one will show that the whole execution plan was to satisfy a query (SELECT statement). In this case, you could determine in which order the operations would have to occur to produce the desired results. With a complicated statement that requires tens of operations to complete, however, you may need to see organized rows from the plan table. Oracle has published a number of variations of a query that shows the relationship between each operation through levels of indentation. The query mentioned previously would be formatted to look something like SELECT STATEMENT TABLE ACCESS BY ROWID INDEX UNIQUE SCAN where the most indented operation is done first, the outermost operation done last, and those in between done according to the amount they're indented. Each level of indentation represents a level of dependency-the outermost operation depending on all the previous levels for it to complete. When two or more operations contribute equally to a parent operation, such as when two sorted sets of table rows are compared in a sort-merge join, they appear under the parent operation at the same level of indentation. Figure 16.2 shows how you can build a tree structure from an execution plan that's formatted in this way. The figure also indicates how you can follow the order of execution from the tree structure if you build it as shown. Figure 16.2 : You can create a tree structure from a formatted plan table query. To help you get started with the EXPLAIN PLAN utility, Listing 16.1 shows a script you should run from SQL*Plus after you create a plan table. It prompts for the statement you want explained, for a statement ID to keep the results separate, and lets you choose whether to delete the resulting rows from the plan table when you're done. Listing 16.1 EXPLAIN.SQL-Script to generate and display a formatted execution plan
http://www.informit.com/content/0789716534/element_016.shtml (13 of 23) [26.05.2000 17:07:06]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

View Code Download this code You can download the EXPLAIN.SQL script from the Web at http://www.Mcp.com/info. You'll be asked to enter an ISBN; enter 0789716534, and then click the Search button to go to the Book Info page for Using Oracle8. Tweak EXPLAIN.SQL I don't expect you to run the EXPLAIN.SQL script as it stands-I rarely do. But I do have a version of it that I use for checking parallel operations, which I call EXPPAR.SQL, and another version, called EXPROWS.SQL, which allows me to concentrate on the number of rows being handled at each step. As you'll see, if you run this script, a number of the fields are truncated to fit the data neatly onscreen. You should make any modifications to this script you would find useful, such as removing unwanted columns from the output and displaying more characters from the columns that interest you. Or you can remove the SUBSTR functions (lines 31 through 34) and allow the data to wrap to multiple rows within the defined column widths. You can see from Table 16.3 that you don't need many characters from the entries in the OPERATIONS and OPTIONS columns to be able to tell them apart (although you may need to keep this book open on your desk until you start to remember them all). When you become familiar with the various execution plans generated by your applications' statements, you can identify useful indexes as opposed to unused ones, spot statements that aren't taking proper advantage of indexes or clustered tables, and locate steps that are causing the most overhead in a statement. You can also gather useful information from examining the execution plans for statements that you're executing with parallel server processes. You want to ensure that most of the steps in the execution path are processed in parallel and, in particular, that you don't have any serial steps interposed between two sets of parallel steps.

The Trace Facility

If you're having problems with application performance that you can't resolve through the interpretation of the execution plans generated by its statements, you may need to track the actual performance of the SQL commands. The trace facility is designed to help you in this task by recording the basic activities performed by a statement as it executes, along with a record of the resources used to perform those activities. The results of a trace are stored in a file that, although it's a readable text file, isn't in the most convenient format for you to find critical information. In addition to the trace option itself, Oracle also offers a tool known as TKPROF, which formats the basic trace files into a report format as well as gives you some control over the contents and organization of the output. If you want to track performance of executing statements, you need to use the trace tool and the TKPROF program.

Creating a Trace File

You can create a trace file in a number of ways. The easiest method to manage is to use SQL*Plus. Setting TIMED_STATISTICS for complete trace data Some statistics that the trace facility can collect involve the length of time taken by certain activities. These values won't be recorded unless the database is running with TIMED_STATISTICS set to TRUE. You can set TIMED_ STATISTICS in your initialization file, with an ALTER SYSTEM file, or in the session where you're planning to run the trace facility. Due to the slight increase in overhead incurred when running with TIMED_STATISTICS enabled, you may want to leave it at its default value of FALSE in your initialization file. However, if your user community will likely use the trace facility and doesn't know about the option to turn on statistics with the ALTER SESSION command, you may need to keep statistics collection turned on for them. Or, if you know when they're likely to be using trace, you can turn it on and off as needed at the database level with the ALTER SYSTEM command.

http://www.informit.com/content/0789716534/element_016.shtml (14 of 23) [26.05.2000 17:07:06]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

Create a trace file 1. Connect to the database in SQL*Plus. 2. Optionally start collecting timing information with the command ALTER SESSION SET TIMED_STATISTICS = TRUE; 3. Start the trace by issuing the command ALTER SESSION SET SQL_TRACE = TRUE; 4. Issue the commands you want to analyze. Unlike using EXPLAIN PLAN, the commands will actually execute, so be careful if you use any DML commands that change production tables. 5. Terminate the trace with the command ALTER SESSION SET SQL_TRACE = FALSE; 6. Disconnect from SQL*Plus and locate the trace file you created. 7. Run TKPROF to format the contents of the trace file. You might find it more convenient to build a script file containing the statements you want to examine and execute that file while tracing is active in your session. This way, you can test the script ahead of time to ensure that you're going to be working with only the statements you intended. You can also use the script to repeat the exercise to check the impact of any changes you may decide to make as a result of your initial tests. If you stay connected to the same Oracle session, trace will continue to use the same trace file no matter how many times you turn tracing on and off. Each time you run the same statement in a single Oracle session, trace accumulates its statistics in a single record for that statement, again whether or not you run it in the same trace session. If you want to compare the before-and-after statistics of a single statement, for example, or a query run with and without a certain index in place, you should disconnect from Oracle before tracing the second execution of the statement. This way, each trace file-one for the first and one for the second session-will include only data for the individual executions of the statement you're investigating. The second way you can initiate the trace facility is to use an Oracle-supplied procedure. This way, you can start the trace on another user's session or start it from within an application program that can execute a PL/SQL block. The procedure, SET_SQL_TRACE_IN_SESSION, is part of the DBMS_SYSTEM package and requires the SID and SERIAL# of the session to be traced supplied as arguments. You can find these values by querying the V$SESSION dynamic performance view. Obtaining access to DBMS_ SYSTEM Not all users may be allowed to execute procedures in the DBMS_SYSTEM package, which is owned by SYS. Other users may need to have the execute privilege on the pack-age granted to them, by SYS or by a user with GRANT OPTION on the package. Non-SYS users will also need to include the schema name, SYS., as a prefix to the pack-age name, or else create a synonym to identify the schema and package. Trace a session for any user 1. Query the V$SESSION table to obtain the SID and SERIAL# for the user's session you need to trace: SELECT sid, serial# FROM v$session WHERE username = 'oracle_username' AND osuser = 'operating_system_userid'; 2. From within SQL*Plus, execute the procedure to start the trace for the selected session: EXECUTE dbms_system.set_sql_trace_in_session(sid,serial#,TRUE) You simply invoke the procedure by name without the EXECUTE command from within a PL/SQL block. 3. Do nothing while the user continues to work. 4. From within SQL*Plus, execute the procedure to stop the trace for the selected session: EXECUTE dbms_system.set_sql_trace_in_session(sid,serial#,FALSE) You simply invoke the procedure by name without the EXECUTE command from within a PL/SQL
http://www.informit.com/content/0789716534/element_016.shtml (15 of 23) [26.05.2000 17:07:06]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

block. The third, and final, method you can use to start tracing is to set an initialization parameter to cause every session to be traced. The parameter, SQL_TRACE, takes a Boolean value-TRUE turns on database-wide tracing, and FALSE (the default value) causes no default tracing. Just as an individual session can perform its own tracing when the database is running with SQL_TRACE set to FALSE, sessions can disable a statistics collection for themselves even when database-wide tracing is active. In either case, the user issues the ALTER SESSION command as shown earlier. Trace only at the database level under controlled conditions Setting SQL_TRACE = TRUE in your initialization file forces Oracle to trace every session that connects to the database. You're advised not to try this on a production database due to the volumes of data that will likely be the result. Every DBA I've talked to who has tried this has never had time to review all the trace files produced. Most of them couldn't even decide which of them might contain useful information. You should plan to trace all data-base sessions only when you're working with a test database and a controlled user community (whether they be users, developers, or even simulation scripts). Even in these situations, make sure that you have the disk space to store the anticipated output.

Formatting a Trace File with TKPROF

The TKPROF program uses the data from a traced session to create a formatted report. The program can also create a script file that will store the trace results in a database table or build a script file containing the session's traced SQL statements. Options for the TKPROF program allow you to sort the report in order of the most resource-intensive statements traced, choosing which resource, or resources, you want to influence the ordering. You can request a formatted execution plan, generated with the EXPLAIN PLAN utility, to be included in the report for each SQL statement. You also can include or exclude information about recursive SQL statements, aggregate or separate the statistics for a single statement executed by different users, and limit the number of traced statements that appear in the report. The latter is most useful when used with a sort option, allowing you to restrict the report to the top few resource-intensive SQL statements. The TKPROF command has two required arguments: the name of the trace file to be processed, and the name of the output file where the formatted report is to be written. Therefore, to accept all defaults, you can type TKPROF filename1 filename2 at your operating system prompt, substituting the trace file for filename1 and the output file for filename2. You don't even need to include an extension for the filenames; TKPROF assumes that .trc will be the extension for the input file and will use .prf as the default extension for the report file. Table 16.4 describes the options for the TKPROF command. To include an option, type its keyword as shown, followed by an equal sign (=) and the required arguments. Table 16.4 Optional arguments for TKPROF Argument: AGGREGATE EXPLAIN Value(s): Description: YES (default) NO Combines statistics for all executions of the command by all users. Separates statistics for the same command by user. user/password Causes an execution plan to be included with each statement in the report, using the named account to run the EXPLAIN PLAN command. schema.table Table used for the EXPLAIN PLAN output if the (default is user EXPLAIN argument is set. Will use an existing table, if it for the schema) exists, deleting any current entries; otherwise, will create the table temporarily, dropping it when the report is completed. Creates a script file to build a table and store results in the filename database. The default file extension is .prf. YES (default) NO Includes recursive SQL in report. Doesn't include recursive SQL in report.

TABLE

INSERT SYS

http://www.informit.com/content/0789716534/element_016.shtml (16 of 23) [26.05.2000 17:07:06]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

SORT

PRINT RECORD

See Table 16.5 (no Determines the order in which SQL statements are listed in default) the report. Order is in descending order of resource use, where the resource is identified by a keyword from Table 16.5. Restricts the report to include only first integer SQL integer statements, based on the SORT order. Creates a script file to execute all non-recursive SQL filename statements in the trace file. The default file extension is .SQL.

Recursive SQL in TKPROF reports Even if you run TKPROF with the option SYS=NO, certain recursive commands will still appear in the report. These commands are associated with the establishment of the trace environment. You may notice errors listed in the report if you use the EXPLAIN option because you may not have per-mission to access all the tables used by these recursive statements. These errors should be of no concern; they simply indicate that the EXPLAIN PLAN command failed when trying to build an execution plan. An example of a TKPROF command that generates a report containing execution plans built by using the SYS plan table, and also builds a script file to re-execute the statements used in the traced session, might look like the following: TKPROF ora00284.trc jan14hr.rpt TABLE = sys.plan_table EXPLAIN = scott/tiger RECORD = jan14hr.sql TKPROF execution plans are real time When you run TKPROF with the EXPLAIN option, the execution plans will be generated as the report is being created. If the trace file was produced some time before you format it, it's possible that the execution plan used by the command when it executed wasn't the same one you'll see in the report. New statistics-generated with the ANALYZE command for cost-based optimization or the creation of deletion of an index, for example-could cause a different execution plan to be used. You're therefore advised to run TKPROF as quickly as possible after creating the trace file if you expect to see an execution plan used, in all likelihood, when the statements were actually processed. Table 16.5 shows the keywords, and their meanings, that you can use with TKPROF's SORT option to organize your report with the most resource-intensive SQL statement first, and the remaining statements in descending order of resource usage. You can use just one of the sort options from Table 16.5, or you use more than one, enclosing your list in a pair of parentheses and separating the options with commas, as in the following: TKPROF ora00284.trc jan14hr.rpt SORT = (EXECPU, EXEDSK, FCHDSK) To generate this report, TKPROF will compute the sum of the values for each statement's CPU usage during its execute phase with the disk reads performed during its execute and fetch phases, and then sort the statements in descending order of the results. Table 16.5 Sort options for TKPROF Option: PRSCNT PRSCPU PRSELA PRSDSK PRSQRY PRSCU PRSMIS EXECNT EXECPU EXEELA Description: Number of times parsed CPU time spent parsing Elapsed time spent parsing Number of physical reads during parse Number of consistent mode block reads during parse Number of current mode block reads during parse Number of library cache misses during parse Number of times executed CPU time spent executing Elapsed time spent executing

http://www.informit.com/content/0789716534/element_016.shtml (17 of 23) [26.05.2000 17:07:06]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

EXEDSK EXEQRY EXECU EXEROW EXEMIS FCHCNT FCHCPU FCHELA FCHDSK FCHQRY FCHCU FCHROW

Number of physical reads during execute Number of consistent mode block reads during execute Number of current mode block reads during execute Number of rows processed during execute Number of library cache misses during execute Number of fetches CPU time spent fetching Elapsed time spent fetching Number of physical reads during fetch Number of consistent mode block reads during fetch Number of current mode block reads during fetch Number of rows fetched

Interpreting Trace Information

The following is a sample report formatted by TKPROF. The user executed only one statement in the traced session. The report includes this statement and its statistics, but other recursive SQL statements have been edited out. Sample Trace Report Formatted by TKPROF *********************************************************** count = number of times OCI procedure was executed cpu = cpu time in seconds executing elapsed = elapsed time in seconds executing disk = number of physical reads of buffers from disk query = number of buffers gotten for consistent read current = number of buffers gotten in current mode (usually for update) rows = number of rows processed by the fetch or execute call *********************************************************** ... *********************************************************** SELECT * FROM patient, doctor WHERE patient.doctor_id = doctor.doctor_id call count cpu elapsed disk query current ------- ---- ------ -------- ----- ------ -------Parse 1 0.00 0.02 0 0 0 Execute 1 0.01 0.10 0 0 0 Fetch 322 0.01 0.03 5 31 3 ------- ---- ------ -------- ----- ------ -------total 324 0.02 0.15 5 31 3 rows ----0 0 322 ----322

Misses in library cache during parse: 1 Optimizer goal: CHOOSE Parsing user id: 18 Rows Execution Plan ------- -------------------------------------------------0 SELECT STATEMENT GOAL: CHOOSE 0 NESTED LOOPS 0 TABLE ACCESS (FULL) OF 'PATIENT' 0 TABLE ACCESS (BY INDEX ROWID) OF 'DOCTOR' 0 INDEX (RANGE SCAN) OF 'SYS_C00551' (NON-UNIQUE)

http://www.informit.com/content/0789716534/element_016.shtml (18 of 23) [26.05.2000 17:07:06]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

*********************************************************** ... *********************************************************** OVERALL TOTALS FOR ALL NON-RECURSIVE STATEMENTS call count ------- ----Parse 4 Execute 5 Fetch 322 ------- ----total 331 cpu elapsed disk query current ----- -------- ----- ------ -------0.00 0.02 0 0 0 0.01 0.10 0 0 0 0.01 0.03 5 31 3 ----- -------- ----- ------ -------0.02 0.15 5 31 3 rows -----0 0 330 -----330

Misses in library cache during parse: 1 OVERALL TOTALS FOR ALL RECURSIVE STATEMENTS call count ------- ----Parse 0 Execute 0 Fetch 0 ------- ----total 0 cpu elapsed disk query current ----- -------- ----- ------ -------0.00 0.00 0 0 0 0.00 0.00 0 0 0 0.00 0.00 0 0 0 ----- -------- ----- ------ -------0.00 0.00 0 0 0 rows -----0 0 0 -----0

Misses in library cache during parse: 0 5 user SQL statements in session. 0 internal SQL statements in session. 5 SQL statements in session. *********************************************************** Trace file: c:\orant\rdbms80\trace\ora00280.trc Trace file compatibility: 7.03.02 Sort options: default 1 5 0 5 5 65 session in tracefile. user SQL statements in trace file. internal SQL statements in trace file. SQL statements in trace file. unique SQL statements in trace file. lines in trace file.

Watch out for disk space limits when using the trace facilities You must be careful before per-forming a trace to check the avail-able disk space on the machine where the trace file will reside. The file could grow tremendously in size because it captures everything you do, even logging off (and back on again). If you then process the file with TKPROF, you need the additional disk space to store the for-matted output. Some key terms from the report that you should recognize are as follows: r parse is the step in the processing of a SQL statement in which the execution plan is developed, along with checks for valid syntax, object definitions, and user authorization. r execute is the step in the processing of INSERT, UPDATE, and DELETE statements in which the data is modified, and in a SELECT statement in which the rows are identified. r fetch is the step in the processing of a query in which rows are retrieved and returned to the application. r count is the number of times a parse, execute, or fetch step was performed on a statement. r cpu is the total amount of CPU time used for the parse, execute, or fetch step of a statement. The time is measured in seconds and reported to the nearest 1/100 second. Processing that completes in less than 1/100 second will be reported as zero. r elapsed is the total amount of elapsed (wall clock) time used for the parse, execute, or fetch step of a statement. The time is measured in seconds and reported to the nearest 1/100 second. Processing that completes in less than 1/100 second will be reported as zero.
http://www.informit.com/content/0789716534/element_016.shtml (19 of 23) [26.05.2000 17:07:06]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8
r

disk is the total number of data blocks read from disk for the parse, execute, or fetch step of a statement. query is the total number of buffers retrieved in consistent mode for the parse, execute, or fetch step of a statement. Consistent buffers are usually used for queries and may contain older copies of records for read consistency purposes. current is the total number of buffers retrieved in current mode for the parse, execute, or fetch step of a statement. Current buffers are usually used for INSERT, UPDATE, and DELETE activities when the most up-to-date version of the data is required. rows is the total number of rows processed by the execute or fetch step of a statement. Any rows processed by a subquery aren't included in this total. internal SQL statements is a SQL statement executed by Oracle in addition to the statement being processed by the user to allow the user statement to complete. For example, a CREATE TABLE command will use recursive calls to reserve space for the initial extent(s), to define indexes for unique constraints, and similar actions. The statistics for these statements are totaled under the heading OVERALL TOTALS FOR ALL RECURSIVE STATEMENTS. library cache misses means that anytime an object definition is required but not available now in the library cache (part of the shared pool in the System Global Area), it's considered to be a library cache miss.

After you master the trace facility and can generate a formatted report, you should understand what it can tell you about your application code. Obviously, if you're running low on a system resource, such as read/write throughput or CPU cycles, you can use the sort option of TKPROF to help you identify the statements that consume most of these resources. Often, by tuning a few of the worst culprits among all the statements used by an application, you can solve your major resource problems. You can also make some determinations about a statement's efficiency by looking at the specific statistics listed by TKPROF. Over time, you should easily be able to spot statements that consume more resources than the norm, or statements that appear to use more resources than similar statements require. Some specific indications of a poor-working statement include the following: Don't jump to conclusions At certain times, the statistics in a trace report may not indicate the real performance of the statements being monitored. Right after you start up the database, for example, you wouldn't see a lot of additional processing if the statement were executed after the contents of the System Global Area had stabilized. Also, running a statement, or a series of statements, in isolation from the normal work load on the system could bias the results- faster throughput because of less contention on table locks, but with more physical disk reads because the work of loading commonly used data into memory isn't being shared with other users running the same, or similar, statements.
r

A large number of blocks being accessed compared to the number of rows being processed. This generally means that tables are being scanned rather than have a usable index to get to the desired rows. Including the EXPLAIN PLAN output helps you determine if indexes are being underutilized. A large number of parse counts, particularly if for the same user. This could mean that a cursor is being closed in the application that might be more usefully left open for reuse. A row count in the execute column of a query's statistics, particularly if there's close to or exactly one row per execution. This indicates that an implicit cursor is being used in PL/SQL for single-row queries rather than an explicit cursor. This can cause additional client/server traffic because the implicit cursor has to send a query probe for what should be a non-existent row to set a return code. Fetches equal, or nearly equal, to the number of rows returned. This is a problem in client/server environments because each fetch requires overhead that could be avoided by fetching the rows in batches.

Certain system-wide tuning problems can also be surmised from the output in a trace file. If the number of disk reads is close to the total number of buffers used (query plus current), for example, it's possible that the database buffer cache isn't large enough. Similarly, if the number of library cache misses is high, your shared pool might be too small.

Using AUTOTRACE
If you are comfortable with SQL*Plus for developing, testing, or tuning your SQL code, you can take advantage of AUTOTRACE. This option causes SQL*Plus to report analytical information after the successful execution of any INSERT, UPDATE, DELETE, or SELECT statement. The information reported

http://www.informit.com/content/0789716534/element_016.shtml (20 of 23) [26.05.2000 17:07:06]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

is derived from the EXPLAIN PLAN utility and the trace utility, although you can control which elements you want to see. This allows you to see similar information interactively that you otherwise have to collect and format as a separate step. It can, therefore, significantly increase your productivity when you need to monitor the behavior of a particular statement or series of statements. Restrictions on AUTOTRACE AUTOTRACE isn't available when FIPS flagging is enabled, or with TRUSTED Oracle. Also, the formatting of your AUTOTRACE report may change if you upgrade your version of Oracle, and it might be influenced by the configuration of the server. You control the behavior of AUTOTRACE with the SET AUTOTRACE SQL*Plus command. By itself, the SET AUTOTRACE command won't change the status of the session, but it will return the full syntax of the command, which looks like the following: SET AUTOT[RACE] OFF | ON | TRACE[ONLY] [EXP[LAIN]] [STAT[ISTICS]] When you choose OFF, AUTOTRACE stops displaying a trace report. If you set it ON, a trace report will be displayed following the standard output produced by each traced statement. The TRACEONLY option will also display a trace report, but it doesn't print the data generated by a query, if any. The EXPLAIN option shows the query execution path by performing an EXPLAIN PLAN command but suppresses the statistical report. STATISTICS, the final option, will display the SQL statement statistics but will suppress the EXPLAIN option output. If you use ON or TRACEONLY with no explicit options, the output defaults to EXPLAIN STATISTICS. You may find the TRACEONLY option to be useful to suppress the display of rows from large queries. If STATISTICS is specified with TRACEONLY, SQL*Plus still fetches the query data from the server even though the query data isn't displayed. Regardless of the options selected, the AUTOTRACE report is printed after the statement is successfully completed. To use the EXPLAIN option, explicitly or by default, you must first create the table PLAN_TABLE in your schema. I recommend using the UTLXPLAN.SQL script to accomplish this to ensure that the version of AUTOTRACE and the table definition are compatible. As mentioned earlier, you can find this script in the admin subdirectory under your ORACLE_HOME directory. To access STATISTICS data, you must have access to several dynamic performance views. The easiest way to handle the necessary privileges-particularly if you'll need to give a number of users access to AUTOTRACE-is to run the PLUSTRCE.SQL script. You can also find this script in the admin subdirectory under your ORACLE_HOME directory. This will create a role called PLUSTRACE and grant the necessary privileges to it. You must run PLUSTRCE.SQL as SYS and grant the PLUSTRACE role to users who will use SET AUTOTRACE. When SQL*Plus produces a STATISTICS report, a second connection to the database is automatically created. This connection is closed when the STATISTICS option is set to OFF, or you log out of SQL*Plus.

Controlling the EXPLAIN Option Output

Each line of the execution plan has a sequential line number. SQL*Plus also displays the line number of the parent operation. Internally, the execution plan consists of seven columns, described in Table 16.6. Table 16.6 Column definitions used by AUTOTRACE for EXPLAIN output Column Name: ID_PLUS_EXP PARENT_ID_PLUS_EXP PLAN_PLUS_EXP Description: Shows the line number of each execution step. Displayed in Position: 1

Shows the relationship between each step and its 2 parent. Shows each step of the execution plan, including 3 the operation and option (from Table 16.3), the object name, and, if using cost-based optimization, the cost and cardinality. Also includes the optimizer choice in the first row. For statements with parallel or remote steps, the bytes value is also included.

http://www.informit.com/content/0789716534/element_016.shtml (21 of 23) [26.05.2000 17:07:06]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

OBJECT_NODE_PLUS_EXP ID_PLUS_EXP OTHER_TAG_PLUS_EXP OTHER_PLUS_EXP

Shows the database links or parallel server processes if they're used. Shows the line number of each parallel or remote execution step. Describes the function of the SQL statement in the OTHER_PLUS EXP column. Shows the text of the query for the parallel server process or remote database.

4 5 6 7

Columns in EXPLAIN portion of the AUTOTRACE output The first four columns-in other words, positions 1 through 3- appear in every execution plan. The last three columns, positions 5 through 7, appear only if the statement involves parallel or remote operations. Although column 4 appears in all execution plans, it's populated only when the same conditions are true that cause columns 5 through 7 to display; in all other cases, it has a value of NULL. You can alter the display of any of these columns with the standard SQL*Plus COLUMN command. For example, to stop the PARENT_ID_PLUS_EXP column from being displayed, enter COLUMN parent_id_plus_exp NOPRINT The default formats can be found in the SQL*Plus site profile (for example, glogin.sql). When you trace a statement in a parallel or distributed query, in general, the cost, cardinality, and bytes at each node represent cumulative results. For example, the cost of a join node accounts for not only the cost of completing the join operations, but also the entire cost of accessing the relations in that join. If any execution plan steps are marked with an asterisk (*), that denotes a parallel or remote operation. Each of these operations is explained in a separate part of the report, using the last of the three columns described in Table 16.6.

The Statistics Option Output

The statistics are recorded by the server when your statement executes and indicate the system resources required to execute your statement. The client referred to in the statistics is SQL*Plus. Net8 refers to the generic process communication between SQL*Plus and the server, whether or not Net8 is installed. You can't change the default format of the statistics report. In Table 16.7, you can find the name and description of any of the reported statistics you may not understand. Table 16.7 Statistics reported by AUTOTRACE Name: Description: SQL Trace Equivalent Statistic: SQL statements executing on behalf of the user's Internal SQL statement statements Database blocks moved into a buffer and used as is Current Database blocks reconstructed in a Blocks read from disk Bytes written to redo log buffer Bytes sent from the database server to the client (SQL*Plus) Bytes sent from the client (SQL*Plus) to the database server Messages sent between the database to/from client server and the client (SQL*Plus) Sorts completed entirely in memory Sorts completed by using temporary segments on disk Rows retrieved (query) or processed (DML) Query Disk N/A N/A N/A N/A N/A N/A Rows (total)

Recursive calls DB block gets Consistent gets Physical reads Redo size Bytes sent via Net8 to client Bytes received via Net8 from client Net8 roundtrips Sort (memory) Sort (disk) Rows processed

http://www.informit.com/content/0789716534/element_016.shtml (22 of 23) [26.05.2000 17:07:07]

informit.com -- Your Brain is Hungry. InformIT - Using Optimizers and the Analytic and Diagnostic Tools From: Using Oracle8

AUTOTRACE statistics and database tuning If many statements have high values for the Sort (disk) statistics, it could mean than the sort space allocated in your initialization file is too small. You may need to modify the parameters that control sort space. Similarly, the redo statistic can help you judge an appropriate size for the redo buffer latch parameters in your initialization file. Both topics are discussed in detail in Chapter 20, "Tuning Your Memory Structures and File Access." SEE ALSO Details on tuning the redo log buffer, More about sort space utilization and balancing memory use and disk access for sorting, You can use the same criteria to judge the efficacy of a statement from the statistics that equate to those discussed for the trace utility. The additional information from the AUTOTRACE statistics can help you judge whether a statement might be using up excessive bandwidth in a client/server environment (the Net8 statistics), or whether the statement is performing too many large sorts. You may need to examine the execution plan to determine whether any of the sorts could be reduced or even removed.

A Sample Session with AUTOTRACE

Listing 16.2 shows a sample SQL*Plus session using AUTOTRACE. The session uses two different queries: one performing a simple table join, the other a join on two tables defined with a non-zero degree of parallelism, causing parallel execution. Listing 16.2 Tracing statements for performance statistics and query execution path View Code Asterisks point to parallel and remote operations Notice the asterisks next to the NESTED LOOPS, TABLE ACCESS, and INDEX operations (execution steps 2, 3, and 4 respectively in lines 88, 90, and 91). These indicate that the operation is described in more detail later on-in this case, in the lines beginning with the word PARALLEL, associated with these same operation numbers (lines 93, 101, and 102). Notice in Listing 16.2 that the results (line 12) of the initial query (line 4) are displayed along with the execution plan (line 31) and statistics (line 40) from AUTOTRACE due to the choice of the ON option only (line 9). In the second query, on the other hand, the query results and the AUTOTRACE statistics aren't shown because of the TRACEONLY and EXPLAIN options (line 74). You also can see that the CREATE commands (lines 56, 61, and 66) don't cause any output to be generated by AUTOTRACE, and that the execution plan (line 83) for the second query (line 78) is generated even though there are no rows in either table involved. In this second query, you can also see asterisks, indicating operations performed in parallel, on steps 2, 3, and 4 in the first section of the execution plan. For each of these marked steps, there's a further explanation in the second part of the execution plan output, using the same step numbers to identify them. Here, the nature of the parallel operation is described in more detail, including an entry corresponding to the OTHER_TAG column in the plan table, and a copy of the statement passed to the parallel server processes. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_016.shtml (23 of 23) [26.05.2000 17:07:07]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Using Constraints to Improve Your Application Performance

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q

Understanding Constraints Constraint Naming Conventions

r r

Using Oracle's Naming Scheme Developing Your Own Naming Schemes

q q

Creating and Managing Constraints General Constraint Options

r r

Statement or Transaction Enforcement Disabling and Enabling Constraints

q q q q q q q q q

Forcing Input with the NOT NULL Constraint Ensuring Distinct Values with Unique Constraints Creating Distinct Rows with Primary Key Constraints Validating Data Against Existing Records with Foreign Key Constraints Defining Business Rules with Check Constraints Including Constraints in a New Table Definition Adding a Constraint to an Existing Table Modifying and Dropping Constraints Using Default Column Values

Understanding Constraints
Constraints allow you to define certain characteristics for columns in a table along with the table definition. In effect, they allow you to encode business rules about the data allowed in a table in addition to the table definition. Such business rules could include ensuring that no two customers are assigned the same ID number, or preventing an order from being taken without a customer ID number. You should plan to use constraints whenever characteristics of a column or group of columns can't be enforced by the chosen datatype. For example, defining a column to hold the minutes portion of a time stamp as NUMBER(2) wouldn't prevent a user from entering a value of 75, even though this wouldn't be a valid number of minutes. You could, however, define a constraint on this MINUTES column to limit the range of values to be between 0 and 59, inclusive. When in effect, the constraint would prevent INSERT and UPDATE statements from creating a record with a value outside the required range in this MINUTES column.

http://www.informit.com/content/0789716534/element_017.shtml (1 of 19) [26.05.2000 17:08:38]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

Constraints provide efficiencies for application developers as well as for database operations: q Simple to code as part of a table definition q Easy to disable and re-enable as needed q Always in place, regardless of tools used to access the table q More or equally efficient execution compared to equivalent SQL statements q Defined in data dictionary for centralized management q Constraint names appear in exception messages q Have to test only when created, not with each application By using constraints, you reduce the work needed to enforce the related business rules. First, you save application developers a lot of work because they won't need to code an additional check to determine whether the provided value is within the acceptable range. In certain cases-such as ensuring that a value in a column is unique-this check could require coding an SQL statement to query the existing table entries. This, in turn, provides the second benefit-fewer SQL statements improve your database efficiency because the database will have less work to do. Constraints can help reduce client/server network traffic In a client/server environment, constraints save a lot of network traffic that would otherwise be needed if the rules were enforced by the application. When an application enforces business rules that require checks against the current data-base contents, it has to issue additional SQL statements. These must be passed across the network and the results returned across the network. This can take several trips, depending on how the SQL is coded and how much data is returned. If on the other hand you define the rules using data-base constraints, all the checking occurs on the database server and any required recursive SQL is generated and executed at the server, causing no network traffic. A third benefit of constraints is that they're part of the definition of the schema to which they belong, and as a result, comprise part of database exports. Naturally, this results in the constraints being passed between database structures when using the Export/Import functions. In some ways, constraints are to relational tables as methods are to objects-both are integral to each other. As with a good object tool, Oracle8 attempts to process the constraint definitions in memory without generating additional, recursive SQL. Constraints don't just define the characteristics of a single column, going beyond the type and length of data defined. They can include multiple columns when necessary. Further, you can define the implicit relationships between tables by using constraints. This way, you can maintain integrity between parent and child tables.

Constraint Naming Conventions

When you create a constraint, you can allow Oracle to name it for you or you can choose your own name. I recommend choosing your own names even though it requires more work for you during the design and development phases. Oracle's naming convention doesn't take into account your table's name or the columns to which the constraint refers. This makes it difficult for you to remember the names of the constraints that you create. The default name is even less useful when it's included in exception messages related to the constraint. Neither the user who receives the message nor the person responsible for analyzing the problem can determine the nature of the problem from an Oracle-generated constraint name.

Using Oracle's Naming Scheme

Every constraint in an Oracle database must have a name that's unique within its schema. If you don't provide one, Oracle will choose a name for you. These default names have the format SYS_Cn, where n is a five-digit integer with leading zeroes. For example, a constraint could have the name SYS_C00512. Constraint names can be difficult to change Before you begin implementing constraints, you should decide if you're going to develop your own naming scheme; if that's the case, you can begin using it right away. Unlike some database objects (such as tables), which you can rename with a simple RENAME command, you have to drop and recreate a constraint in order to change its name. As your database grows, even this option can become difficult because some constraints have other constraints dependent on them. Such dependencies can prevent you from dropping key constraints.

http://www.informit.com/content/0789716534/element_017.shtml (2 of 19) [26.05.2000 17:08:38]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

Suppose that you have a constraint on the MINUTES column of a table that allowed only valid minute values (0 through 59), and the system had given this constraint the name SYS_C00512. A user attempting to enter a value of 75 into this column would receive an error message naming the constraint that they violated. Imagine how useful the user would find that error message, stating only that constraint SYS_C000512 had been violated. How would the user, or anyone in a support position, determine the nature of the error from that message? Along the same lines, suppose that you decided to store the fraction of an hour in the minutes column rather than the actual number of minutes-30 minutes would now be stored as 50, representing 50 percent of an hour. To save a value representing 52 minutes, you would now need to store 86. To allow this value, you would have to disable the constraint restricting the column to a high value of 59. You can find the constraints placed on a table through the view DBA_CONSTRAINTS (or USER_CONSTRAINTS). Looking at the entries in this table, though, it would be difficult for you to determine that constraint SYS_C000512 is the one you need to remove. Another situation in which you may need to quickly identify a constraint's source is when you're dealing with constraint dependencies. It's very unlikely that you would remember which Oracle-generated name was given to the primary key constraint on a table you created a couple of months ago, but you may need to know when creating another table so that you can verify the constraint's definition. Similarly, if you were to change the status of that primary key constraint, you might be told that your newer constraint was dependent on its current status. Would you know to which constraint the message was referring if it gave you only the name SYS_C00512?

Developing Your Own Naming Schemes

You could adopt a number of possible naming schemes for identifying constraints. Most of them contain three elements in the name: table name, column name, and constraint type. Of course, this may not always be possible if you use long names for either (or both) table and column names. The constraint name is limited to the 30-character maximum set for all Oracle names. If your table name is already 25 characters long, this leaves only five characters for both column name and constraint type. You may therefore need to define a standard set of abbreviations for your table and column names. The reason to use all three elements should be obvious from the preceding section. If you want to drop or modify a constraint on the values allowed in a MINUTES column in a table, you would like to be able to reference the constraint name directly in the SQL statements needed to effect the change. The most complete naming convention would identify the table name, the column name (MINUTES), and the type of constraint. As schema objects, your constraints must have names that are unique within their schema. If you used only table name and constraint type as a naming convention, you would limit yourself to just one constraint of each type in any table, or else you would have to differentiate the constraints with some other naming element. Many people gravitate to a numbering scheme when faced with this type of problem, but if you were to use this approach, you're likely to end up with constraint names such as ACTION_UQ_1 and ACTION_UQ_2. This is almost as useless for quick identification as is the infamous constraint named SYS_C00512. There are only five constraint types; the sections devoted to each type appear later in this chapter. It's quite simple to assign a meaningful two-character string to identify each type. For example, UQ was used in the preceding paragraph. UQ is a widely used abbreviation for the unique constraint. Although this is jumping the gun, Table 17.1 shows the types of constraints, along with a suggested two-character abbreviation for use in your constraint names. Table 17.1 Constraint types and suggested abbreviations for naming Constraint Type: PRIMARY KEY UNIQUE NOT NULL FOREIGN KEY CHECK Suggested Abbreviation: PK UQ NN FK CK

Alternatives for identifying constraint types

http://www.informit.com/content/0789716534/element_017.shtml (3 of 19) [26.05.2000 17:08:38]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

You may already have noticed that the five constraint types can be individually identified by the single initial letter of their types. If you use long table and column names, you may need to reduce the constraint-type indicator in constraint names to this single letter. However, you may find yourself using initial letters in other names, such as when naming a column in one table that's related to a column in another table-for example, the column in the LAB_RESULTS table that contains the patient's ID number from the PATIENTS table. You might name this column P_ID. Obviously, such single letters could become confused with a constraint type in some circumstances. The two-character, constraint-type indicators suggested in Table 17.1 are unlikely to be chosen for any other use (hence the UQ rather than UN for unique), thus avoiding possible ambiguity. Besides deciding to use the three-part constraint names, you also need to determine in which order these parts appear in the name. You should be able to find five distinct abbreviations for the constraint type that won't be confused with either table or column names, so this isn't a major concern. However, it's possible for a table and a column (in the same or a different table) to have identical names. You could mistake constraints based on an ambiguous name unless you have a standard to ensure that these two elements are always included in the constraint name in the same order. Another point to consider is the use of constraints that cover multiple columns. Consider a constraint name such as PATIENT_TREATMENT_ID_UQ. This could be a unique constraint on the ID column of the PATIENT_TREATMENT table, or on the combined TREATMENT and ID columns of the PATIENT table. You reduce this ambiguity if you separate the table and column names with the constraint-type abbreviation. Thus, the preceding example would become PATIENT_TREATMENT_UQ_ID or PATIENT_UQ_TREATMENT_ID, depending on the name of the base table.

Creating and Managing Constraints

Unlike other database objects that have their own CREATE, ALTER, and DROP commands, constraints are so closely tied to the tables they control that they're manipulated by the commands associated with those tables. As you may remember from Chapter 7 "Adding Segments for Tables," the syntax for creating and altering tables is quite complex in itself. The concepts and the syntax for dealing with constraints is also rather complicated. When you put the two together, the resulting syntax diagram becomes almost impossible to read. Consequently, as you look at each of the options associated with constraints, just the keywords or key phrases required by the syntax to manage that option are introduced. When you're comfortable with these concepts, this chapter ends with a look at the overall constraint syntax and you can see how it fits into the various table management commands. Table or column name first in constraint name? Whether you start the constraint name with the table or the column name(s) is your choice. The dictionary views DBA_/ALL_/USER_ CONSTRAINTS don't contain the column names, so if you want to query these tables and report the column names in sorted order, the column name(s) must come first. This approach doesn't help if you have a constraint that covers multiple columns because there will be only a single constraint name; hence, only the first columns in the constraint name will appear in the desired position in the query's sorted out-put. You can, of course, use the DBA_/ALL_/USER_ CONS_COLUMNS views alone or joined to the *_CONSTRAINTS tables to find the column names and order them in the result set. In the next few sections you examine the options you can apply to constraints that in turn help you and your users make the most effective use of the options. Some options, such as the capability to temporarily turn off constraint checking, apply to all constraint types; other options are specific to one type of constraint. The latter are covered in individual sections devoted to each constraint type. Begin by looking at constraint options that include the options available to all constraints. An overall syntax diagram for a constraint definition looks like this: [CONSTRAINT constraint_name] constraint_type [constraint_options] You were briefly introduced to constraint types in Table 17.1, and we will look at each type in detail later in this chapter. The options you look at first, as mentioned, are those valid for all constraint types.

http://www.informit.com/content/0789716534/element_017.shtml (4 of 19) [26.05.2000 17:08:38]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

General Constraint Options

In the following sections you look at the various options available to you regarding the enforcement of constraints. These options apply to all types of constraints; they are identified and given values in the segment of the constraint definition labeled constraint_options in the constraint syntax diagram. When a constraint is defined, it stays associated with the table to which it applies. However, it may not always be convenient to have the constraint enforced all the time. You have various options to determine when new records, or even updated records, will be checked for validity against the relevant constraints. In some cases you may want to check as soon as a statement is processed; on the other hand, you may want to wait until a whole work batch is complete before identifying constraint violations. CONSTRAINT keyword and constraint name are optional Leaving these terms out of a constraint definition will cause Oracle to generate a constraint name, in the form of SYS_Cn, as discussed earlier. Including these two elements names the constraint with the name you provide. To determine when and how to force a constraint check when each individual statement or transaction executes, you should read the next section, which discusses immediate and deferred constraint checking. To look at reasons for turning off constraints for an arbitrary period of time, read the "Disabling and Enabling Constraints" section.

Statement or Transaction Enforcement

When a constraint is in place, every row affected by an SQL statement is examined for constraint violations, and an exception is raised if at least one row remains in violation when the statement completes. This behavior allowsinterim violations to exist for the duration of the statement, but not beyond. An interim violation is one that, although breaking the rules of constraint while processing is under way, is revolved by the time the processing completes. As a simple example, consider a table containing a company's department information. The departments are numbered one by one, but due to a reorganization, the company needs to make the numbers multiples of 10; that allows new subdepartments to be added in the gaps. A partial query of the DEPARTMENTS table's original contents and the UPDATE statement needed to make the required change are shown in this SQL*Plus session: SQL> REM Examine current values in DEPARTMENT_ID column of SQL> REM DEPARTMENTS table SQL> SELECT department_id 2 FROM departments 3 WHERE department_id <= 20 4 ORDER BY department_id / DEPARTMENT_ID 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 20 rows selected.
http://www.informit.com/content/0789716534/element_017.shtml (5 of 19) [26.05.2000 17:08:38]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

REM Multiply each DEPARTMENT_ID by 10 to REM change increment from 1 to 10 SQL> UPDATE department 2 SET department_id = department_id * 10 / 65 rows updated. REM Check the work SQL> SELECT department_id 2 FROM departments 3 WHERE department_id <= 20 4 ORDER BY department_id / DEPARTMENT_ID 10 20 SQL> At some point during the execution of this UPDATE statement, the old DEPARTMENT_ID 1 became 10, the old DEPARTMENT_ID 2 became 20, and so on. If the DEPARTMENTS table had a constraint to prevent duplicate DEPARTMENT_ID values at the time these changes were made, there would have been constraint violations, because the original department 1 and the original department 10 both would have a value of 10 in their DEPARTMENT_ID columns. This would also have been the case for departments 2 and 20, departments 3 and 30, and on up to departments 6 and 60. If Oracle caused the statement to fail due to these anomalies, it would be very difficult for users to make these perfectly valid, albeit infrequent, changes. To avoid this, Oracle marks the duplicates as interim violations and then, when the statement has completed all its changes, it checks them again to see if they remain as violations. By this time, these changes would have included changing the original departments 10, 20, 30 to 100, 200, 300, hence removing duplicate values for 10, 20, 30. The statement can therefore complete without any problems. By default, Oracle checks interim constraint violations at the end of each statement. If one or more are found, the statement fails with an exception and all the changes are rolled back. In some cases, you may want to delay the constraint checking because you need to combine the effects of two or more statements to create consistent records. Suppose, based on the preceding example, that employees are associated with each DEPARTMENT_ID value in the DEPARTMENTS table. Oracle can enforce a rule (by using a constraint type that's discussed later) that all employees must have a valid department number as part of their records. If we make the preceding change to the DEPARTMENTS table, any employee registered to departments other than 10, 20, 30, 40, 50, or 60 would have an invalid record. To correct this, we would also have to change the employee records. This would take a second statement-thus, the default behavior is of no use because the first statement, the one changing the DEPARTMENTS table, would fail. Changing the employee records first wouldn't help either because we would need to change employees in department 65 (for example) to be in department 650, and such a department number doesn't exist in the unchanged DEPARTMENTS table. Deferred constraints allow you to code cascading updates When two tables contain related information-such as a PATIENTS table that contains a field for the patient's doctor's ID, stored in the DOCTORS table-deleting or updating records can be problematic. For example, to change a doctor's ID value in the DOCTORS table would leave the related patient records without a valid doctor. The update to the value in the DOCTORS table needs to be cascaded to the appropriate records in the PATIENTS table. The ANSI standard doesn't allow such cascading updates when a constraint is used to enforce the relationship between the two tables. By using deferred constraints, how-ever, Oracle will let you make changes to both tables within a single transaction before applying constraint checking. You can use this capability to update the doctor's ID in both tables via two separate statements, thereby coding your own cascading update. To complete the required changes to both tables-the DEPARTMENTS table and the one with the employee records-we need to defer the constraint checking until the department numbers are changed in both. This is done by using what are known as deferred constraints. If you expect to use deferred constraints for any reason, you must understand what options are available. There are basically two approaches to deferring constraints: one requiring the application user to defer any required constraints at the time the transaction begins, and one allowing the constraint to be deferred automatically in all transactions. Within the constraint definition options, these activities are controlled with the following keywords: q DEFERRABLE

http://www.informit.com/content/0789716534/element_017.shtml (6 of 19) [26.05.2000 17:08:38]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8
q q q

NOT DEFERRABLE INITIALLY IMMEDIATE INITIALLY DEFERRED

The DEFERRABLE keyword determines whether the constraint can be deferred at the time the transaction begins. A constraint can't be deferred by an application when defined as NOT DEFERRRABLE. When defined as DEFERRABLE, the following SQL command will allow any interim violations of the named constraint to remain until the transaction is committed, regardless of when the statement that caused the violations occurred within the transaction: SET CONSTRAINT constraint_name DEFERRED You can make multiple constraints deferrable with the SET CONSTRAINT command by naming them in a comma-separated list. Alternatively, you can issue the following command to defer checking all the constraints encountered in the transaction until it completes: SET CONSTRAINT ALL DEFERRED If you use a list of constraints in the command, they must all be defined as deferrable-otherwise the command will fail. If you use the keyword ALL, only the deferrable constraints, if any, will become deferrable for that transaction. The SET CONSTRAINT command is no longer in force as soon as the transaction completes with a COMMIT or a ROLLBACK. The INITIALLY keyword sets a deferrable constraint's default behavior. If set to INITIALLY DEFERRED, a constraint is automatically deferred within any transaction that encounters it. You don't need to issue the SET CONSTRAINT command to defer checking on such a constraint until the end of a transaction; it will be done that way anyway. By default, however, a constraint is INITIALLY IMMEDIATE; this means that all interim violations are checked at the end of each statement and the statement will fail if any are found. Because the INITIALLY keyword is valid only for constraints already defined as DEFERRABLE, the SET CONSTRAINT command can be used to override DEFERRED or IMMEDIATE. We have already seen that the following statement will defer checking the named constraint until the end of the transaction: SET CONSTRAINT constraint_name DEFERRED Similarly, this command will cause the named constraint to be checked at the end of each statement in which it's invoked during the course of the transaction, even if it's defined as INITIALLY DEFERRED: SET CONSTRAINT constraint_name IMMEDIATE In either case, the scope of the SET CONSTRAINT command is a single transaction; unless you reissue it at the start of your next transaction, the default behavior will apply to all constraints again. One final note on deferred transactions: During such a transaction, you can issue the following command to see if any interim violations now exist: SET TRANSACTION ALL IMMEDIATE Handling multiple interim violations If more than one constraint has interim violations when you issue the command SET TRANSACTION ALL IMMEDIATE, only one will be reported. Therefore, you can't be sure if you only have one or if you have multiple violations. If you want to address all violations, you have to correct the one that's reported by SET TRANSACTION ALL IMMEDIATE and then reissue the command to look for further violations. If they do, the command will return an error message about the violation, such as ORA-00001: unique constraint (SYS_C00315) violated or ORA-02292: integrity constraint (SYS_C00894) violated - child record found.

Disabling and Enabling Constraints

In the preceding section you saw how to defer constraint checking until the end of a transaction. Sometimes you might find it useful to disable a constraint for an even longer period. Suppose you're building a new medical database in which you have two tables, such as the DOCTORS and the PATIENTS tables, with dependent information. You might want to use a constraint to ensure that every patient is assigned to a doctor. (The section
http://www.informit.com/content/0789716534/element_017.shtml (7 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

"Validating Data Against Existing Records with Foreign Key Constraints" covers this type of constraint, known as a foreign key or referential integrity constraint later in this chapter.) If the constraint to enforce this logical relationship between doctors and patients were enforced, you would have to load all doctors' records first and then all patients' records. This would prevent you from running multiple load programs, some storing patient data and some doctor data, to shorten the overall load time. The data loading problem could become even more complex if the patient table included a field for the financially responsible party and forced a referential integrity check between patients and their financial representatives. If the dependent patient's record were loaded before the responsible party's record, there would be a constraint violation and the load would fail. Even deferred constraint checks couldn't help us if the records for the two related patient records were being stored through two different load programs; each program would be managing just its own transactions. To help you with such situations, Oracle allows you to temporarily turn off a constraint and then to restart its enforcement later. These actions are known as constraint disabling and enabling, respectively. Disabling Constraints In general, there are a number of reasons for turning off constraint checking when performing large data loads, whether into a brand new database or adding to existing data. Rather than make you drop the entire constraint definition, Oracle lets you disable a constraint. The constraint remains defined in the data dictionary while it's disabled, but it's not enforced. You can re-enable a disabled constraint whenever you're ready. At this time, Oracle will check for any violations, and if it finds one or more, will return an exception message and leave the constraint disabled. When you enable a constraint and expect (or already know) that there will at least violation, you can ask Oracle to save information about which row or rows are causing the exception in a special table. By using data from this table, you can extract the non-conforming rows into a temporary table for later inspection and resolution. With this done, you can then re-enable the constraint on the remaining rows, thus protecting the table from any further violations. First look at the various ways you can disable a constraint. The methods include the following: q Creating the constraint in a disabled state q Disabling an enabled constraint by using the constraint type q Disabling an enabled constraint by using the constraint name q Explicitly disabling a dependent constraint By default, all constraints are created in the enabled state. If you include the keyword DISABLE in the same clause as you define it, however, your constraint will be defined but disabled. To enforce such a constraint, you would have to enable it at a later time. Certain types of constraints can be disabled by identifying just the constraint type, others by identifying the constraint type and the column or columns on which the constraint is defined. Either type, plus any other type of constraint, can be disabled if you know its name. Because only one primary key constraint can be defined on a table, the syntax to disable it can be as simple as DISABLE PRIMARY KEY. The syntax that disables a unique constraint is almost as simple. In this case, the key phrase is DISABLE UNIQUE (column_name), which identifies the name of the column on which the constraint is defined. In some cases, the unique constraint will span multiple columns. In such cases, the parentheses need to contain not just one column name, but a list of the relevant column names separated by commas. Primary-key and unique constraints, as well as any other constraint, can be disabled by naming them in the DISABLE CONSTRAINT constraint_name clause. Here, only one constraint can be named in the phrase. You would need to either enter multiple disable phrases or issue multiple SQL commands in order to disable more than one constraint on a table. In some cases, a constraint may have another constraint that depends on its existence. You can't disable a constraint with such a dependency unless you also disable the dependent constraint. Although you can issue a separate command to disable the dependent constraint first, you can also use the DISABLE clause's CASCADE option to disable any dependent constraints at the same time you disable the parent constraint. DISABLE PRIMARY KEY CASCADE is an example of this type of statement. DISABLECASCADE has no corresponding ENABLE CASCADE

http://www.informit.com/content/0789716534/element_017.shtml (8 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

If you have to disable a dependent constraint to disable a parent constraint, you need to keep track of which constraints you disable, so that you can re-enable them after you re-enable the parent constraint. This shouldn't pose a problem if you issue explicit SQL commands to disable the dependent constraints-you can perform the disable commands via a script and simply use a modification of the script to re-enable them. You won't have a record of what, if anything, is being disabled if you rely on the DISABLE command's CASCADE option to disable the related constraints. Although you can query the data dictionary to find which constraints are now disabled, it won't show you which of these were disabled as a result of any particular cascade action. You can't assume that you should re-enable them all after you re-enable the primary constraint- some of them may have been disabled for other reasons. One final issue that you should understand about disabling constraints concerns primary-key and unique constraints. Oracle uses an index to enforce these. If the index had to be built when the constraint was enabled, the disabling action will also drop the index. In addition to stopping the integrity checking, the loss of the index may reduce the performance of statements that could normally use the index to reduce the number of rows accessed. Enabling Constraints By default, constraints are enabled when they're first defined. As you've just seen, they can be created in a disabled state or disabled at a later time. You need to use one form of the ENABLE constraint clause whenever you need to make a disabled constraint active. In its simplest form, the ENABLE phrase resembles the DISABLE phrase discussed in the preceding section. That is, it can be in one of the following forms: q ENABLE PRIMARY KEY q ENABLE UNIQUE (column_name,[column_name][,...]) ENABLE CONSTRAINT constraint_name These commands shouldn't have any problems executing if either no changes have been made to the table since the constraint was disabled or if the table is empty. On the other hand, if there are new or changed rows in the table, the constraint may not be re-enabled due to records that violate the constraint. If this is the case, you have two options: q Reactivate the constraint with the ENABLE NOVALIDATE phrase rather than just the ENABLE phrase (which actually includes an implied VALIDATE keyword that you could code if you wanted). This will cause all further changes to be subject to the constraint but won't examine the existing data. If you use this approach, you may need to find a way to go back and correct the invalid rows at a later time. q Identify the rows and deal with them, thus allowing the constraint to be enabled successfully. To do this, you need to add the phrase EXCEPTIONS INTO table_name to the ENABLE phrase. The table_name can be either just a table name or a schema and table name. Whatever table you use, it must be formatted in a specific way for the command to work. Building a table to hold constraint exception information The easiest way to build an exceptions table is to run the Oracle-supplied script UTLEXCPT.SQL, which creates an EXCEPTIONS table in your own schema. The enable phrase that allows you to use this table while enabling a primary key is ENABLE PRIMARY KEY EXCEPTIONS INTO exceptions. If you want to use a table with a different name or in a different schema, substitute the appropriate table reference in the ENABLE phrase. No matter which table you use, it must defined with exactly the same columns and datatypes as the default table in UTLEXCPT.SQL. If you use an exceptions table, any rows that violate a constraint when you try to enable it are identified in this table. Each such row has an entry in the EXCEPTIONS table showing its rowid, the name of the constraint that the row violates, and the name and owner of the table to which the row belongs. After you identify the rows that violate a constraint, you need to either update them to fix the problem or delete them, depending on the nature of the problem and the application's requirements. In many cases, you may want to defer the corrections until a later time but still want to enable the constraint to avoid further potential violations. The following script contains a series of commands that attempt to activate a unique constraint in the DOCTORS table and then moves any problem rows out of DOCTORS and into a temporary table, FOR_FIXING, where they can be processed later: ALTER TABLE doctors ENABLE UNIQUE (suite_number) EXCEPTIONS INTO exceptions
http://www.informit.com/content/0789716534/element_017.shtml (9 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

/ CREATE TABLE for_fixing AS SELECT * FROM doctors WHERE rowid IN (SELECT row_id FROM exceptions) / DELETE doctors WHERE rowid IN (SELECT row_id FROM exceptions) / TRUNCATE TABLE exceptions / ALTER TABLE doctors ENABLE UNIQUE (suite_number) / Unless additional invalid changes are made to the DOCTORS table between the time you start this script and the time the final command is executed, the constraint should be enabled when the script completes. When you or your users have updated the FOR_FIXING table created by the script, you can try to put the rows back into the original table with the following command: INSERT INTO doctors SELECT * FROM for_fixing; You may need to address one other option when enabling certain constraints. The unique constraint and the primary key constraint require an index to help enforce them. The index may not exist while a constraint is disabled, in which case it will need to be built when the constraint is enabled. You need to include the USING INDEX phrase as part of the ENABLE clause if you don't want the new index to use your default tablespace, or you want to override one or more of the default storage or space utilization parameters. Should the index already exist, the USING INDEX clause will be ignored. The syntax for the USING INDEX phrase is as follows: USING INDEX [PCTFREE integer] [INITRANS integer] [MAXTRANS integer] [STORAGE ( [INITIAL integer [K|M] ] [NEXT integer [K|M] ] [PCTINCREASE integer] [MINEXTENTS integer] [MAXEXTENTS integer] [FREELISTS integer] [FREELIST GROUPS integer] ) ] TABLESPACE tablespace_name NOSORT [NO[LOGGING]] The terms used in this phrase are a subset of the options for the CREATE INDEX command, which you can find described in detail in Chapter 8 "Adding Segments for Different Types of Indexes." SEE ALSO A complete description of the options when creating an index,

Forcing Input with the NOT NULL Constraint

You use the NOT NULL constraint when your table contains a column that's required to hold a value for every single row in the table. If you're familiar with entity-relationship models, an attribute at the mandatory end of a relationship would always have to contain a value. In a business environment, it might be a requirement that every order in the ORDERS table contain a valid method of payment, such as a credit-card number, a bank account number for customers paying by check, or a flag to indicate that cash has been received. A NOT NULL constraint can't be defined on multiple columns, but you can use a NOT NULL constraint on as many different columns in a single table as you need. To add a NOT NULL constraint, use the clause CONSTRAINT constraint_name NOT NULL. Remember that if you don't mind Oracle naming your constraint for you, you can omit the keyword CONSTRAINT and the constraint's name. You can also include
http://www.informit.com/content/0789716534/element_017.shtml (10 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

any of the deferrable and validation options discussed earlier if you're assigning the constraint as part of a CREATE TABLE command, or if you're adding a new column to an existing table. However, you can't use these options when you're adding a NOT NULL constraint to an existing column. NOT NULL constraints in the data dictionary If you're looking for a NOT NULL constraint in the data dictionary tables, such as DBA_ CONSTRAINTS, you may wonder why you don't see any constraints of type N, or some other obvious code letter for these constraints. The reason is that Oracle internally enforces NOT NULL requirements with a check constraint, such as those you can build yourself (as described a little later). Therefore, the NOT NULL constraints are tagged with the code letter C (the abbreviation for check type constraints) in the CONSTRAINT_ TYPE columns of the data dictionary views.

Ensuring Distinct Values with Unique Constraints

You may need one or more columns in a table to contain no duplicate data. A table containing data being collected in real time may need each record to contain a distinct time stamp, for example, so that a user can tell the exact order in which the data was gathered. You may have to ensure that no two employees share the same userid in the company's email system when you add new records to the EMPLOYEES table. A unique constraint provides a simple method to enforce such a requirement. A unique constraint can be defined on a single column or on a collection of columns. In the latter case, it's known as a composite unique constraint. If a single column is constrained with a unique key, you can still store multiple records with a NULL value in that column. With a composite unique constraint, two records are considered duplicates if they contain identical values in every column covered by the constraint. The same value can appear in the same column of many records without violating the constraint if at least one column in each record has a different value. NULLs in columns of a composite unique constraint are ignored when comparing two different records, so uniqueness has to be provided by the values in the non-NULL columns. Table 17.2 shows pairs of records from a two-column composite unique constraint and identifies which are considered duplicates. Table 17.2 Unique and non-unique characteristics of composite constraints Column 1: 14 12 12 12 12 13 12 12 12 124 NULL NULL Column 2: 15 13 13 14 NULL NULL 13 13 NULL NULL NULL NULL Uniqueness: Unique Unique Unique Duplicates Duplicates Unique

As with other constraint types, you can add unique constraints by using your own name for them, or allow Oracle to name them for you. You also can make them deferrable or not, with the default behavior for deferrable constraints dependent on whether the user begins a transaction as deferrable. Similarly, the constraint can be created in an enabled or a disabled mode. (See the preceding sections for details of these characteristics.) The full syntax of the clause for creating unique constraints is as follows: [CONSTRAINT constraint_name] UNIQUE [column_name[, column_name[...]]] [deferred_clause] [enabled clause] [exceptions_clause] [index_clause] CONSTRAINT constraint_name applies your name to the constraint. Omitting it will let the name be an Oracle-supplied name.

http://www.informit.com/content/0789716534/element_017.shtml (11 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8
q q

q q

UNIQUE is the required keyword for a unique constraint. deferred_clause contains any required DEFERRABLE and initial phrases and values. (See the detailed syntax in the earlier section "Statement or Transaction Enforcement.") enabled_clause contains either ENABLED or DISABLED. exceptions_clause identifies the table where invalid rows can be identified, as discussed in the "Enabling Constraints" section. index_clause has the following syntax: USING INDEX [PCTFREE integer] [INITRANS integer] [MAXTRANS integer] [STORAGE ( [INITIAL integer [K|M] ] [NEXT integer [K|M] ] [PCTINCREASE integer] [MINEXTENTS integer] [MAXEXTENTS integer] [FREELISTS integer] [FREELIST GROUPS integer] ) ] TABLESPACE tablespace_name NOSORT [[NO]LOGGING]

The terms in the USING INDEX clause are identical to those involved when building a new index with the CREATE INDEX command; see Chapter 8 where they're described in detail. Without a USING INDEX clause, the required index will be created in the same tablespace as the table if it's a new column and a new constraint; if it's a constraint being enabled it's built in the user's default tablespace. Of course, if there's already an index on the column(s) covered by the unique constraint, you shouldn't include the USING INDEX clause-otherwise the statement will fail. Indexes created automatically through the addition of a unique constraint are also dropped automatically when the constraint is dropped or disabled. Oracle names these indexes with the same name as the constraint itself, whether it's an Oracle- or a user-supplied constraint name. If the index exists on the table before the constraint is created, it can't be dropped unless the constraint is dropped or disabled. You can't name your constraint with the same name as the index in such cases. Indexes for unique constraints You should use your own index-rather than rely on the index created as part of the constraint definition-when you need the performance benefits offered by the index. This way, the index will always be avail-able, regardless of the status of the constraint. Depending on your needs and the nature of the data, you can use a unique or non-unique index to enforce a unique constraint. A non-unique index might be useful if you have needs for duplicate values during processing that occurs while the constraint is disabled. SEE ALSO A complete description of the options when creating an index,

Creating Distinct Rows with Primary Key Constraints

A relational database table can't distinguish between two rows that have exactly the same values in all their columns. It's therefore advisable to create every table with a primary key constraint. This type of constraint guarantees that every row has some unique value that makes it distinguishable from every other row. Although there can be many columns in some tables that fulfill this requirement, only one primary key constraint is allowed per table. A primary key constraint has the characteristics of the NOT NULL and the unique constraints. However, if the primary key is a composite constraint (a composite primary key constraint)-that is, defined across more than one column-the NOT NULL characteristic is applied to each individual column. Therefore, you can't store a row with a NULL value in a primary key column. This is different from composite unique constraints, which do allow NULLs in one or more columns. Other than the constraint type keyword, the syntax for a primary key constraint is identical to that for a unique

http://www.informit.com/content/0789716534/element_017.shtml (12 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

constraint. The syntax is shown here, but you're referred to the preceding section, which discusses unique constraints, for the details of the various clauses: [CONSTRAINT constraint_name] PRIMARY KEY [column_name [,column_name[...]]] [deferred_clause] [enabled clause] [exceptions_clause] [index_clause] An index is used just as for unique constraints to enforce a primary key's unique qualities. An index enforcing a primary key behaves in exactly the same way as one for a unique constraint, including what happens if it pre-exists or if it's created automatically; see the previous discussion on unique constraints for information about primary key indexes.

Validating Data Against Existing Records with Foreign Key Constraints

Earlier in this chapter, we discussed two tables that contained records with a dependency between them. In case you skipped that section, the tables were for a medical facility and one, DOCTORS, contained information about the physicians who practice there; the second was a PATIENTS table with information that included the ID number of the patient's doctor. To ensure that every patient is assigned to a doctor in the practice, the DOCTOR_ID column in the PATIENTS table needs to be validated against the ID column in the DOCTORS table. This can be achieved with a foreign key constraint, also known as a referential integrity constraint. You create foreign key constraints by identifying the column (or columns if it's a composite foreign key) in the parent table that contains the data against which new records are to be checked. In the example you would put a foreign key constraint on the PATIENTS table's DOCTOR_ID column. This constraint would identify the DOCTORS table's ID column as the column containing the data against which patient records would have to be checked. Unlike primary key or unique constraints, an index isn't required to support a foreign key constraint. However, without an index on the foreign key, some SQL statements performed against the parent table may cause locks to be placed on the table with the foreign key. These locks can prevent one or more users from accessing the table concurrently and might slow overall processing as a result. It's therefore recommended that you build an index on any foreign key constraints that you define. This will reduce the number of locks on the table and cause less restrictive locks to be placed on the index instead. Before you try to add a foreign key constraint to a table, you need to confirm that the parent table has a primary key or a unique constraint defined on the columns you need your foreign key to reference. Oracle relies on this constraint to maintain a valid relationship between the two tables. If the constraint on the parent table is disabled or dropped, the foreign key also has to be disabled or dropped. A CASCADE option for these commands causes the appropriate action on the foreign key to occur. You also need to determine what happens to records in your table if the related record in the parent table is deleted. You have the option of preventing the DELETE from occurring if you have a child record still in place, or you can elect to allow the DELETE to delete any records with the same value in the foreign key column(s). The syntax for a foreign key constraint is as follows: [CONSTRAINT constraint_name] [FOREIGN KEY (column_name[, column_name[...]]) ] REFERENCES [schema_name.]table_name [(column_name[, column_name[...]] ) ] [ON DELETE CASCADE] [deferred_clause] [enabled clause] [exceptions_clause] The CONSTRAINT clause optionally assigns the given name to the constraint. The FOREIGN KEY clause is required only if the constraint is being defined outside the scope of the column definition (a column constraint versus a table constraint, as discussed later in this chapter). When required, it also must name all the columns covered by the constraint. The keyword REFERENCES is required and identifies the parent table.

q q

http://www.informit.com/content/0789716534/element_017.shtml (13 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8
q

q q

The schema_name is required if the parent table isn't in the same schema as the table to which the foreign key constraint is being assigned. The table_name is the name of the parent table. The column_name, or series of column_names, are the columns in the parent table that hold the values being referenced; they're required only if they don't comprise the parent table's primary key. ON DELETE CASCADE is an optional clause that allows the deletion of records from the table if the referenced parent record is deleted. the deferred_clause, enabled clause, and exceptions_clause perform the identical actions as described in detail for unique constraints.

To explain the action of the ON DELETE CASCADE option, consider the example of the DOCTORS and PATIENTS tables. The PATIENTS table has a foreign key constraint on the DOCTOR_ID column, which references the ID column in DOCTORS. Suppose that a doctor has an ID of 22 in the DOCTORS table and that the PATIENTS table has a number of records for patients of doctor number 22. If this doctor decides to leave the practice, the record for this doctor should be deleted from the DOCTORS table. If this were allowed without any further action the patient records with DOCTOR_ID equal to 22 would violate the foreign key constraint. By default, Oracle would prevent the deletion of the doctor record because of this resulting constraint violation. However, if the constraint were defined with the ON DELETE CASCADE option, the deletion of the record for doctor 22 from the DOCTORS table would also cause the records with DOCTOR_ID equal to 22 to be deleted from the PATIENTS table. These cascaded deletes occur automatically and without any feedback to the user that any patients are being deleted. Don't allow cascading deletions if they might not always be required Be careful when allowing cascaded deletes on a foreign key and provide this option only when it's always going to be valid for the child records to be removed automatically. In this example, the business rules are likely to stipulate that patients be assigned to different doctors when a doctor leaves. If the ON DELETE CASCADE were in place, an enthusiastic employee could, without even knowing it, delete all the patient records for the departing doctor before such reassignments were completed. If the patients also owned laboratory results and similar records stored in additional tables with foreign key constraints allowing cascading deletes, these would also be lost as a result of the unintended action. Although you may not need to use them very often, you can also define self-referencing constraints. These are foreign key constraints where the constrained column is in the same table as the referenced column, or (to put it another way) where the parent table and the child table are one and the same table. A parts and assemblies table may use such a constraint to ensure that every component listed for an assembly is itself a valid component or assembly as verified by another entry in the table.

Defining Business Rules with Check Constraints

Check constraints allow you to write your own conditions to determine whether the value in a column is valid. The conditions must evaluate to true or false, but they can include comparisons between two different columns in the same row as well as comparisons of column values to constants or functions. There are many uses for check constraints. Some that we've encountered include the following: q Requiring that the value in a gender column be M or F q Ensuring that birth date is at least 18 years less than hire date q Checking that both or neither of two specific columns contain NULLs q Avoiding storing a negative quantity-on-hand value Although they're powerful, check constraints may not be able to define all your needs. For example, they can't compare values in two different rows in the same table, nor check values in a different table. The former restriction also means that you can't include an aggregate function, such as MIN or AVG, in a check constraint because they implicitly obtain data from multiple rows. In addition, check constraints can't include subqueries or functions that return values based on something outside the record being constrained. These restrictions are all related to the issue that if the constraint were valid at one point in time, a change to some other record could invalidate it. For example, the AVG function is disallowed because any new record in a table could change the average value of a numeric column. Similarly, a subquery that retrieves records from the table could easily return different results after the new row for which the constraint is being checked is stored. Functions such as SYSDATE and USERID and pseudocolumns such as ROWNUM will be different depending on when the row is stored, who stores it, or where in the table it's located.

http://www.informit.com/content/0789716534/element_017.shtml (14 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

For complex rules, you can add as many check constraints to a column as you need to guarantee valid data. If a business rule becomes too complicated, you might not be able to use a check constraint or any other type of constraint to enforce it. You'll need to add a trigger to the table, or possibly handle the validation in the application code. Although not as efficient as a constraint, these techniques are required for certain types of integrity enforcement. The check constraint syntax is quite simple compared with other constraints, and consists of the standard naming, deferred, and enabling options as well as the CHECK keyword and the related condition. The condition is placed inside parentheses, resulting in the following structure: [CONSTRAINT constraint_name] CHECK (condition) [deferred_clause] [enabled clause] [exceptions_clause] All components of this syntax, except the CHECK clause, were discussed in detail earlier in the section on unique constraints, so this information isn't repeated here. Instead, look at the condition clause by examining some examples. Table 17.3 shows the business rules listed at the beginning of this section and a check condition that could perform the required validation. Table 17.3 Sample business rules and related check constraints Rule: Require that the value in a gender column be M or F Ensure that birth date is at least 18 years less than hire date Check that both or neither of two specific columns contain NULLs Avoid storing a negative quantity-on-hand value Check Clause Syntax: CHECK (gender IN ('M','F') CHECK (hire_date - birth_date> 18 * 365.25) CHECK ((col1 IS NULL AND col2 IS NULL) OR (col1 IS NOT NULL AND col2 IS NOT NULL)) CHECK (quantity_on_hand >= 0)

Including Constraints in a New Table Definition

When a new table is created, you can include any required constraints as part of the CREATE TABLE command. For single-column constraints, you can include the constraint definition as part of the column clause or add the definitions following all the column clauses. Constraints defined with their columns are called column constraints by Oracle, whereas constraints defined independently of their columns are called table constraints. This distinction is made only because the constraints' syntax varies slightly in some cases, depending on whether the constraint is defined at column or table level. Any constraint will work in exactly the same way, regardless of which type of definition was used to create it. Some constraints can be defined only by using one or the other constraint type. Composite constraints, for example, must be defined as table constraints because they aren't part of any single column definition. The general syntax for defining a table with constraints is as follows: CREATE TABLE table_name ( column_name column_datatype [(length)] [column_constraint [column_constraint [...]]], column_name column_datatype [(length)] [column_constraint [column_constraint [...]]], ... column_name column_datatype [(length)] [column_constraint [column_constraint [...]]], table_constraint [, table_constraint [...]] )

Column constraints versus table constraints

http://www.informit.com/content/0789716534/element_017.shtml (15 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

Unless the nature of the constraint- a multi-column constraint, for example, which must be defined at the table (not column) level-dictates it, you can use a column and a table constraint interchangeably. The difference in how the constraint is defined is simply a matter of style, not functionality. If you're using scripts to build your tables, you may prefer to define the constraints along with the related columns whenever possible, so that all the pertinent information about each column is in one place in the script. Alternatively, you may want to put all the constraint definitions in one place in the script, using table constraints to achieve this. Again, you can't do this for all constraints because a NOT NULL constraint can be defined only at the column level. Chapter 7includes details of the column_name, column_type, and length definitions, as well as other clauses (not shown here) relating to the table's extents, space utilization, and similar characteristics. We want to examine only the constraint clauses here. You can see from the syntax that a column definition can include one, many, or no constraint clauses. If there are constraint definitions, they all precede the comma that separates one column definition from the next. Table constraints follow the last column definition and are separated from it and from each other by commas. The other major difference in syntax between the two types of constraint definitions is that, because it's implied, a column constraint doesn't need to include the column name as part of its definition. Naturally, a table constraint must include the column name or names to which it applies because there's no syntactical link between the definition and any specific column definition. There are some specific restrictions regarding constraint definitions when creating a table: q A NOT NULL constraint must be defined as a column constraint. q A composite constraint must be defined as a table constraint. q Only a table-level foreign key constraint requires the FOREIGN KEY clause. q Constraints enforced through the existence of an index (primary keys and unique constraints) can't violate the limits of an index-namely, they can't include more than 32 columns, and the total key length must be less than half the size of an Oracle block. As an example of a table using different constraints, look over the following script file, which contains a CREATE TABLE command: CREATE TABLE orders id NUMBER(10) CONSTRAINT orders_pk_id PRIMARY KEY, order_date DATE CONSTRAINT orders_nn_date NOT NULL, customer_id NUMBER(10) CONSTRAINT orders_nn_customer NOT NULL CONSTRAINT orders_fk_customer REFERENCES CUSTOMERS (id), status CHAR(5) CONSTRAINT orders_ck_status CHECK (status IN ('NEW','SHPD','PRTL','CANC','HOLD')), total_order NUMBER(20,2), shipped_value NUMBER(20,2), unshipped NUMBER(20,2), payment_type VARCHAR2(15), credit_rating CHAR(4) CONSTRAINT orders_ck_credit_rating CHECK (credit_rating IN ('EXCL','GOOD','FAIR','POOR',UNKN')), sales_rep NUMBER(10), sales_region NUMBER(10), CONSTRAINT orders_ck_ship_total CHECK (total_order=shipped_value+unshipped), CONSTRAINT orders_ck_payment_rating CHECK (payment_type IS NOT NULL OR credit_rating IS NOT NULL), CONSTRAINT orders_fk_sales_rep_region FOREIGN KEY (sales_rep, sales_region) REFERENCES employees (id, region) ) /
http://www.informit.com/content/0789716534/element_017.shtml (16 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

Adding a Constraint to an Existing Table

You need to use the ALTER TABLE command to add a constraint to a column in an existing table. You need to use the table constraint definition for all constraint types (except NOT NULL). The command is as follows: ALTER TABLE table_name ADD ( table_constraint [, table_constraint [...]] ) As you can see you can add more than one constraint with a single command; and, as with the CREATE TABLE command, each constraint clause can include any valid option for the constraint type. These options are described in detail in the earlier sections of this chapter. If you need to make a column NOT NULL, you need to use the ALTER TABLE command with the MODIFY keyword. The syntax for this command is as follows: ALTER TABLE table_name MODIFY ( column_name column_constraint [, column_constraint [...]] ) column_constraint is a NOT NULL constraint with any optional clauses you want to use. Again, as with adding new constraints, you can use a single command to add multiple NOT NULL constraints to a table. If you're adding one or more new columns to an existing table, you also use the ALTER TABLE command but now include the column definition. This definition may also include one or more constraints, or you can include table constraints in the statement following the new column definitions.

Modifying and Dropping Constraints

You might need to modify the constraint definitions you've created for a table. These modifications could include enabling a constraint, disabling or dropping a constraint, or changing a NOT NULL column to allow NULLs. All these changes require the use of the ALTER TABLE command. The earlier section on enabling and disabling constraints looks at the details of the enable and disable clauses, so they're not repeated here. To perform these actions, you need to embed these clauses in the ALTER TABLE command: ALTER TABLE table_name enable_clause ALTER TABLE table_name disable_clause You can include multiple enable or disable clauses in a single ALTER TABLE command, and even use both types of clauses in a single statement. However, you can name only one constraint in each clause. The following script contains the code necessary to redefine the index used for the primary key constraint defined on the ORDERS table's ID column. It shows the use of both the DISABLE and ENABLE options of the ALTER TABLE command.

http://www.informit.com/content/0789716534/element_017.shtml (17 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

To drop the index associated with the primary key constraint, you need to disable (or drop) the constraint Use the keywords to disable the constraint; because there's only one primary key per table The index will be rebuilt when you enable the primary key (again using the key-words) Include this clause to over-ride default index creation Use this clause to override index creation in your default tablespace Include other space utilization parameters if needed You don't need to leave free space if the index is on a column with an ever-increasing value Use whatever options are appropriate for your index You may want to include this in case an invalid record was added while the constraint was disabled You need to name your table if you don't use the default created by ULTEXCPT.SQL

Dropping a constraint is similar to disabling one. You have the same options to identify the constraint: by name or by the keywords PRIMARY KEY and UNIQUE, the latter with the list-included columns. The CASCADE option is also available should you wantto force any foreign key constraints that depend on the constraint being dropped concurrently. As with the disable action, you can't drop a constraint if there's a dependent constraint. The syntax for dropping a constraint follows. (Refer to the section on disabling constraints if you need clarification on any of the included terms; they perform identical duties in either command.) ALTER TABLE table_name DROP [PRIMARY KEY] [UNIQUE (column_name[, column_name [...] ] ) ] [CONSTRAINT constraint_name] As with the enable and disable clauses, you can include only one constraint in an ALTER TABLE...DROP command. Again, the NOT NULL constraint doesn't completely conform to the syntactical rules of the other constraint types. Although you can use all the preceding commands to enable, disable, or drop a NOT NULL constraint, you can also use an ALTER TABLE...MODIFY statement to switch a column between allowing and disallowing NULLs. The statement has the following form: ALTER TABLE table_name MODIFY (column_name [CONSTRAINT not_null_constraint] [NOT] NULL [,column_name [CONSTRAINT not_null_constraint] [NOT] NULL] [,...] ) Tablespaces containing tables with dependent constraints You may have to drop a tablespace that contains tables with foreign key dependencies. You have to use the INCLUDING CONTENTS clause in order to drop a tablespace that still contains tables. The DROP TABLESPACE command will fail if any of the tables that would be dropped when you issue this statement are parent tables for foreign key constraints belonging to tables in a different tablespace. To override this, you can add the CASCADE CONSTRAINTS clause to the DROP TABLESPACE... INCLUDING CONTENTS command. As with the use of this option on individual tables, you may want to defer using it until you confirm that you won't affect any applications by indiscriminately dropping all the dependent constraints. As you can see, this command allows you to alter the NULL enforcement for a single column or multiple columns in a single command. You can even include column changes from NULL to NOT NULL and from NOT NULL to NULL within the same statement, although you can't change the same column twice within one command. The command also gives you the option of using the constraint name (if there is one), but works just as well if you don't include it. One final issue you need to consider when using constraints: Your command will fail if you try to drop a table
http://www.informit.com/content/0789716534/element_017.shtml (18 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Constraints to Improve Your Application Performance From: Using Oracle8

that's being referenced by at least one foreign key constraint in a different table. The exception message will indicate that the primary key or a unique key in the table is required by the foreign key reference. To drop the table, therefore, you must drop the foreign key constraint. You can do this by manually dropping the foreign key constraint before you attempt to drop the referenced table. You can also elect to include the keyword CASCADE CONSTRAINTS clause in your DROP TABLE command in order to remove all dependent constraints. However, because it does this automatically without letting you know how many or which constraints are dropped, it isn't necessarily a wise option to use. It may behoove you to check why any dependent constraints exist before dropping the parent table.

Using Default Column Values

Oracle provides a mechanism-although it's not strictly a constraint-to supply a default value for a column when the initial INSERT statement fails to provide a value. The default can be a constant or an expression, provided it doesn't include a reference to another column or to CURRVAL, NEXTVAL, LEVEL, or ROWNUM. A common use of default values is to store the current date and time obtained through the SYSDATE function whenever a record is added to a table. The following column definition shows this type of default: time_stamp DATE DEFAULT sysdate

A default column value can be changed at any time with an UPDATE command, so you would also need to add a constraint to the column if you want to restrict its allowable values when a record is created. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_017.shtml (19 of 19) [26.05.2000 17:08:39]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Using Indexes, Clusters, Caching, and Sorting Effectively

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

Tuning Indexes
r r r r

When to Use B*Tree Indexes When to Use Bitmap Indexes When to Use Reverse-Key Indexes When to Use Index-Organized Tables Evaluating Index Usage Watch for SQL Statements That Don't Use Indexes Creating and Managing Index Clusters Creating and Managing Hash Clusters Restrictions on Clustered Tables Evaluating Cluster Usage Setting Up a Multiple-Buffer Pool Caching Objects with the CACHE Attribute Evaluating Caching Effectiveness Understanding Sort Behavior Optimizing Sort Operations Setting Sort-Related Parameters Managing Temporary Segments

The data is frequently retrieved via the primary key.

r r r r r r

Caching Data
r r r

Tuning Sorts
r r r r

Tuning Indexes
Indexes are the first in line of means to reduce disk I/O and to provide a short access path to the desired data. Until version 7.2, Oracle used to provide only b*tree indexes. The task for selecting an index was relatively simple then: have one or not? Several indexing options are available with Oracle8, and now it's challenging to select an optimal indexing strategy. Indexes are transparent to users Indexes are transparent to the end-user application. Oracle automatically maintains the indexes during a DML operation on the table. Oracle Optimizer chooses the suitable index during a SQL SELECT operation.

http://www.informit.com/content/0789716534/element_018.shtml (1 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8

Chapter 8 "Adding Segments for Different Types of Indexes," discusses indexes in more detail; the following sections give you a better idea of how to use indexes effectively with your databases.

When to Use B*Tree Indexes

Balanced tree (b*tree) indexes have been part of the Oracle RDBMS since its inception. It's still the most commonly used indexing mechanism. You should create a b*tree index on a table column if it meets any of the following criteria: q SQL queries frequently use the column in the where clause, and the amount of data returned by the query is very low compared to the table data. Consider a sales order-entry table, Orders, which has the following columns: order# order_date part# Quantity customer# sales_person Increase data selectivity Use a concatenated index (that is, an index on multiple table columns) if these columns are involved in any of these situations. You can create a concatenated index to increase the selectivity of the data by using indexed columns. However, ensure that leading columns from a concatenated index are used frequently in an application's SQL statements.
q

If you frequently retrieve orders with a query containing order# in the where clause (as shown in the following SQL command), you must create an index on this table on the Order# column. select * from orders where order# = <order number>; Similarly, other columns on the table can be indexed if they're also used in the queries' where clause and return very few rows. However, imagine a query like the following: select * from orders where qty = '1' If a large proportion of your orders are with this quantity, an index on this column won't be useful because the query will return a large number of rows. Using a full table scan in that situation is more appropriate. The column is frequently used to join multiple tables in queries. Again consider the order-entry table mentioned in the preceding example, along with a parts_master table that contains columns describing attributes for parts such as part#, Description, price, and so on. To prepare the invoice for an order, you might have to execute a SQL statement very similar to the following: select ord.order#, ord.customer#, ord.part#, parts.Description, ord.Qty, parts.Price from orders ord, parts_master parts where ord.part# = parts.part# ; Indexes on orders and parts_master tables on the part# column are recommended in order for this query to run fast. The column is defined as a foreign key to enforce a referential integrity constraint. Frequent update operations involve this column on the parent table. The presence of an index on the child table will avoid a table-level shared lock on it, which Oracle acquires while updating the parent key column in the absence of the index. Oracle automatically creates unique indexes to enforce primary-key and unique-key integrity constraints. You must also create an index on columns in tables that are referred as parent columns in a foreign-key constraints relationship with other tables.

q q

When to Use Bitmap Indexes

Bitmap indexes, first introduced in Oracle 7.3, are very frequently used in data warehouse environments. Creating a bitmap index is recommended in the following situations: q The column or the set of columns frequently appear in the where clause of SQL statements.

http://www.informit.com/content/0789716534/element_018.shtml (2 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8
q

q q

The column has fewer distinct values as compared to the total number of rows in the table. The ratio of distinct values to the total number of values is known as cardinality. Thus, the column has a low cardinality value. The column is used frequently to perform joins between multiple tables. Updates to the column are infrequent.

When to Use Reverse-Key Indexes

A reverse-key index is very similar to a regular b*tree index, except that the byte order for the indexed column is reversed. Consider a table containing the orders where the primary key, order numbers, is generated in increasing sequence order. Because all the entered orders are nearby, all the new values will go to the same index leaf block, thus causing multiple concurrent inserts into the same block. This is especially a problem in an Oracle parallel server environment in which that block will be pinged among the instances for insertion of new values. If the byte order for the order number is switched now, the numerically adjacent order number will go to different leaf blocks, and pinging is greatly reduced. Reverse-key indexes are recommended in the following situations: q The indexed column is a serially increasing number. q There are concurrent INSERTs and DELETEs in the table. q The application is used in an Oracle parallel server environment. q Queries don't require index range scans-that is, the queries involving the indexed column that use clauses such as between, greater than, lesser than, and like aren't likely to be used.

When to Use Index-Organized Tables

An index-organized table (or index-only table) is very similar to a b*tree index. If your table data is very static and will invariably be retrieved via an indexed key, an index-only table can be a better alternative to combining a regular table and its associated index-it will save the additional I/O to retrieve the additional column from the table data after the index lookup. You can't create another index on an index-only table, however. Use index-organized tables in the following situations:

The data is frequently retrieved via the primary key.

q q

A set of columns is most frequently selected during the retrieval. No other table column will be used in the where clause of SQL statements; hence, the index isn't needed on any other column. Full table scans are very infrequent.

Consider a table containing a Social Security number, name, and address information. If queries similar to the following are going to use the table extensively, it is advantageous to create the table as an index-organized table: Select SSN, Name, Address from cust_info where ssn = <SSN> ;

Evaluating Index Usage

The only direct way to confirm the usage of an index is to look at your application code; do this to determine whether the indexed column is appearing in the where clause of a SQL statement. You should reconfirm the index usage by looking at the query's execution plan. However, using an index and avoiding a full table scan doesn't guarantee the optimal indexing strategy. You should further examine the possibility of creating a more selective concatenated index on a table if multiple columns from the same table are appearing in the SQL statement's where clause. Decisions could be simple in an OLTP environment where a transaction deals with a relatively low volume of data indexing. During batch processing, during report generation, or in a DSS environment, however, the decision between various indexing strategies and full table scans performed in parallel could be tricky and might be greatly influenced by machine resources, database layout, the nature of the query, and so on. A full table scan can be faster than using an index

http://www.informit.com/content/0789716534/element_018.shtml (3 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8

Under favorable conditions, a full table scan using the Oracle parallel query option might be faster than using an existing index. This is especially true while joining multiple tables. At what point a full table scan becomes faster than an index retrieval depends on many factors, including the available CPU capacity and a table's physical layout. The following command sequence shows how to use the auto-trace utility to quickly get the execution plan of a SQL statement and confirm the index usage: Confirm the presence of needed indexes with na execution plan Look at your time-consuming queries' execution plan to see whether performance can be improved by adding an index or by increasing the selectivity of an existing index (via adding another column to it). In an OLTP system where resource usage by a transaction isn't significant, however, the overall system can reduce the I/O marginally, because every transaction is executed numerous times. SQL> set autotrace on ; SQL> select * from item where i_id=1000 ; ITEM_ID ITEM_NAME ITEM_PRICE ---------------------------------- ---------ITEM_DATA -------------------------------------------------1000 FAbLm1A84sVcZXgkJbZvSVe 7590 HTXSxYlPUMW5HGc5umArHcJofKDlwiOXPN Execution Plan ----------------------------------------------------------

Index used to retrieve data from ITEM table To measure the overall effectiveness with which an Oracle instance is using indexes, calculate the Index Use Ratio as follows: Index Use Ratio = 'table fetch by rowid'/ (table fetch by rowid + table scan rows gotten) table fetch by rowid and table scan rows gotten are the statistics from the dynamic performance view V$SYSSTAT. A value of 90 percent or higher for this ratio is recommended. A lower value might be acceptable in a data-warehousing or decision-support system where the full table scans are frequently used. Index usage doesn't come without a cost. Before creating an index, consider the following factors along with the advantages you'll gain by the presence of an index: q The DML operations involving the indexed column will also need to update the index along with the table, thus consuming more resources than used to update the table. q Indexes require additional disk space. q Creating an index on a large table could be time-consuming.

Watch for SQL Statements That Don't Use Indexes

Several SQL queries don't use indexes: q SQL queries that perform calculations or functions on the indexed column. For example, the following SQL statement won't use an index on the name column: Select name, address from persons where upper(name)='JOHN';

http://www.informit.com/content/0789716534/element_018.shtml (4 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8
q

SQL statements that use NOT when testing an indexed column. The following SQL statement won't use the index present on the city column: Select name, street_address, city from persons where city not in ('BOSTON','NEWYORK'); SQL statements that use an IS NULL or IS NOT NULL clause. SQL statements that perform internal conversion on the indexed column. In a part_master table where part# is defined as a character column, the following query won't use the index on the part# column: select * from part_master where part# = 999999; Instead, use this SQL statement: select * from part_master where part# = '999999';

q q

A cluster is a group of tables. If the application SQL statements frequently join two or more tables, you can improve performance by clustering the tables. Oracle stores the clustered tables in the same data blocks. This reduces the amount of I/O required during the retrieval, because the rows needed from multiple tables are available in the same block. Similarly, you can group multiple table rows with the same value for a non-unique column together if you know that these rows will be processed together. This type of cluster-where a key column is used to group multiple rows-is known as an index cluster. The key column used to group the rows together is the cluster key. When clusters hurt performance Using clusters for frequently updated or deleted tables affects performance adversely. Another type of cluster that Oracle offers is hash clusters-single-table clusters where rows of a table with the same hash value are stored together to enable fast access. These are discussed in detail in a later section. Consider a telephone billing system where the calls table stores information about each customer's usage. If billing rules indicate that all calls made by a customer in a particular month be billed together, you can store these calls in a cluster whose cluster key could be concatenated for the columns phone_number, year, and month. Figure 18.1 shows the difference between a normal table and a single-table cluster. Figure 18.2 shows the data storage in a three-table cluster. Figure 18.1 : Index clusters store the data with the same index key together. Figure 18.2 : Index clusters store the data from multiple tables with the same cluster key together.

Creating and Managing Index Clusters

Create an index cluster and its tables (general steps) 1. Create the cluster. 2. Create the cluster index. 3. Create the tables. 4. Insert data and perform DML/query operations. When to use index clusters If two or more tables are related via referential integrity and frequently used with joins in the SQL operations, create an index cluster for these tables. If multiple rows from a table are frequently fetched together with a non-unique column, create a single-table index cluster with that column as the cluster key to improve performance. Creating an Index Cluster Use the following command to create a single-table cluster for the previously discussed calls table: SQL> create cluster C_phone_calls ( phone_no number(10), year number(4), month number(2))

http://www.informit.com/content/0789716534/element_018.shtml (5 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8

pctused 80 pctfree 5 tablespace users storage (initial 10M Next 10M); Creating a Cluster-Key Index Oracle Server requires an index on the cluster key before it allows any DML performed against the tables in the index cluster. You can't perform any DML operations until you create the index. To locate a row with a given cluster key, Oracle first looks in the index and reads the corresponding rowid. This rowid is, in turn, used to retrieve the table data for the clustered table(s). The following command creates the cluster-key index I_phone_calls on the c_phone_calls cluster: Cluster-key indexes can't be unique Oracle doesn't allow cluster-key indexes to be unique. You can create indexes on a clustered table's other columns. These indexes are maintained independently of the cluster index. SQL> create index I_phone_calls on cluster c_phone_calls pctfree 10 tablespace user storage (initial 1M next 1M); Creating Tables in an Index Cluster You can create a table within a cluster after creating the cluster. A cluster-key index can be created before or after the tables within the cluster. However, you can't insert any rows in a cluster's tables until you create the cluster-key index. The following CREATE TABLE command creates the table phones_calls within the cluster c_phone_calls: SQL> create table phones_calls ( phone_no number(10), year number(4), month number(2),day number(2), duration number(3), .......... ) cluster C_phone_calls (phone_no number(10), year number(4), month number(2));

Creating and Managing Hash Clusters

Hash clusters are single-table clusters in which rows with the same hash-key values are stored together. Oracle uses a mathematical hash function to choose the location of a row within the cluster. A cluster key and the data block in which it will be stored are directly related. The space used for a cluster needs to be allocated during cluster creation. Thus, the size and the number of rows should be accurately known. When to use hash clusters Use hash clusters for lookup tables whose data is static and is retrieved mostly by using their primary key. Consider an item table containing information about 100,000 items and having the following structure: item_id item_name item_price item_data number(6,0), varchar2(24), number(5,0), varchar2(50)

Also assume that the item_id is unique and the average row length is 75 bytes. If a 3,750-byte space is available within each Oracle data block, each data block will accommodate 45 rows (3,750 divided by 75) per block. The total storage space needed for this cluster will be 2,223 data blocks (100,000 divided by 45). Oracle rounds it off to a certain higher number and might allocate a few more blocks. The data stored in the cluster will look similar to Figure 18.3.

http://www.informit.com/content/0789716534/element_018.shtml (6 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8

Figure 18.3 : Hash clusters preallocate the storage location for a row based on the hash-key value. When Oracle needs to store or retrieve a row with an item_id, it simply applies the hash function to the given item_id to get its block number. In the example, the hash function is mod(45); to retrieve data for item_id 10576, Oracle looks in data block 236 (10,576 divided by 45), needing to read only one disk block. Use the following SQL statement to create a hash cluster for this data: create cluster item_cluster ( item_id number(6,0)

create table item ( item_id number(6,0), item_name varchar2(24), item_price number(5,0), item_data varchar2(50) ) cluster item_cluster(item_id); Total number of cluster keys to be stored Cluster-key column, based on data most frequently retrieved from table Space (in bytes) used to store data for single-cluster key Carefully choose the size amount If space used by the rows associated with a cluster key isn't predictable, exercise caution while using clusters. If space usage frequently exceeds allocated space, chaining will take place and will result in wasted disk space and increased I/O.

Restrictions on Clustered Tables

Clustered tables have some restrictions: q A direct path load can't be used to load tables in a cluster. q Storage allocation is defined at the cluster level; thus, it can't be defined for each individual table. q Clusters can't be partitioned. q The cluster index can't be unique. q Clustered tables can't contain long, long raw, or LOB object datatypes.

Evaluating Cluster Usage

Oracle Server automatically maintains the clusters once they're created. Oracle will also use the underlying hash or the index cluster key for data retrieval whenever the cluster key is used in a SQL statement's where clause. The following example uses the autotrace utility to see the SQL query's execution plan. The execution plan shows that the table is accessed via HASH of ITEM, thus using the hash cluster key to retrieve the specified data. SQL> set autotrace on ; SQL> select item_id,i_name from item where item_id=1000; ITEM_ID I_NAME ---------- -----------------------http://www.informit.com/content/0789716534/element_018.shtml (7 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8

1000 FAbLm1A84sVcZXgkJbZvSVe Execution Plan ---------------------------------------------------------0 SELECT STATEMENT Optimizer=CHOOSE 1 0 TABLE ACCESS (HASH) OF 'ITEM'

Caching Data
Caching is the most common approach for improving retrieval rates of large random data. The effectiveness of caching depends on the size of the cache memory available and the access pattern of the data. Oracle uses a significant part of the Shared Global Area (SGA) as the buffer cache. Its size can be specified by the DB_BLOCK_BUFFERS initialization parameter in terms of total number of Oracle data blocks, and it's generally referred to as the buffer cache or buffer pool. By default, Oracle uses the Least Recently Used (LRU) algorithm for maintaining the buffer cache, in which the least recently used blocks are used first. When an Oracle user process needs a data block it proceeds as follows: 1. It looks in the buffer cache for the desired block. 2. It finds a free buffer in the buffer cache and reserves it. 3. If it doesn't find a free buffer after searching a predetermined number of buffers, it signals the DBWR. The DBWR, in turn, writes the modified data blocks to the disk. The buffers written to the disk are available for reuse. 4. The process reads the block from the disk in the buffer cache. Figure 18.4 depicts these steps in the form of a flowchart. Figure 18.4 : An Oracle process follows this algorithm to access a data block. The effectiveness of caching depends on the access pattern for the data. The access pattern may vary vastly from object to object, however. To improve caching effectiveness, Oracle8 offers a multiple-buffer pool and a CACHE attribute for the objects.

Setting Up a Multiple-Buffer Pool

Oracle8 lets you divide the buffer pool into three different pools: q KEEP As the name suggests, buffers in this pool are kept as long as possible. Frequently accessed objects that contain a small number of blocks should be assigned to this pool. q RECYCLE Buffers in this pool are recycled after they're used. You should assign to this buffer pool very large, randomly accessed objects whose blocks won't likely be accessed again very soon. q DEFAULT Data block buffers remaining after assigning the specified buffers to the KEEP and RECYCLE pools are assigned to this pool. Objects that haven't been allocated to any other buffer pool are cached in this pool. By default, Oracle8 enables only one buffer pool-that is, the default buffer pool and all the buffers allocated to it. To assign buffers to the KEEP and RECYCLE buffers, use the following initialization parameters:

Assigns 10,000 total buffers to the pool Assigns 10 LRU latches Assigns 1,000 buffers and 2 latches to the KEEP pool Assigns 3,000 buffers and 2 latches to the RECYCLE pool Figure 18.5 shows the buffer pool allocation specified by these parameters.

http://www.informit.com/content/0789716534/element_018.shtml (8 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8

Figure 18.5 : Oracle8 can divide a buffer pool in three sections. Assigning Objects to Buffer Pools By default, all objects use the DEFAULT buffer pool. To cache the objects in other pools, you need to create the object with the desired buffer pool attribute or alter the object's buffer pool attribute by using the ALTER command. ALTER and CREATE for the table, partition, index, cluster, snapshot, and snapshot log support the buffer pool attribute in their storage clause. You can assign different buffer pools to different partitions of a partitioned object. Use the following command to create and assign the example item table to the KEEP buffer pool: item_id number(6,0), item_name varchar2(24), item_price number(5,0), item_data varchar2(50) ) storage (initial 1M next 1M buffer_pool keep); If the table already exists, use the following command to assign it to the KEEP buffer pool: Alter table item storage (buffer_pool keep);

Caching Objects with the CACHE Attribute

In addition to grouping the objects in the KEEP, RECYCLE, and DEFAULT buffer pools according to their access patterns, you can also specify the CACHE attribute for very frequently accessed tables; this makes it so that buffers for these tables are read at the most recently used end of the LRU during their full table scans. Choosing objects to cache The caching strategy for a particular object depends on its access pattern and the available buffer cache. If object is a frequently used reference table, assign it to the KEEP buffer pool. On the other hand, assign large, randomly accessed table to the RECYCLE buffer pool. The proper caching strategy should factor the available memory, disk I/and overall response time for transactions. By default, Oracle always keeps the read blocks at the least recently used end of the LRU list during a full table scan. You can specify the CACHE attribute for the table with the CREATE or ALTER command to keep the read blocks at the most recently used end of the LRU chain. Use this command to alter the item table to use the cache attribute: Alter table item cache; If you don't want to permanently change a table's CACHE attribute but still want to read it at the most recently used end of the LRU list during the current full table scan (due to special processing requirements), use the CACHE hint in the SELECT statement. Similarly, you can use the NOCACHE hint to disable a table's CACHE attribute during the current full table scan.

Evaluating Caching Effectiveness

The effectiveness of a buffer cache is measured in terms of the cache hit ratio or the buffer's hit ratio, defined as the percentage of times a desired data block was found in the buffer cache. You can calculate the buffer hit ratio from the statistics available in the V$SYSSTAT dynamic performance view as follows: Cache hit ratio = 100*(1- 'physical reads'/ 'Logical reads') Logical reads is further derived as the following: Logical Reads = 'db block gets' + 'consistent gets' These statistics from the V$SYSSTAT view give you the combined statistics for the buffer cache as a whole. If you've defined KEEP and RECYCLE buffer pools, it's strongly recommended that you evaluate the hit ratios for them separately. The V$BUFFER_POOL dynamic performance view gives physical reads, block gets, and consistent gets statistics for each buffer pool.

http://www.informit.com/content/0789716534/element_018.shtml (9 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8

Tuning Sorts
Sorting is the rearranging of a random collection of data sets into an ordered collection-sorting the data in order. You need to sort the data to make it more presentable and usable. Sorting data consumes memory and CPU resources. The time taken and resources consumed to sort are proportional to the amount of data sorted. If the volume of the data to be sorted is more than the available memory, the sort operations must use disks to store intermediate sort results (further slowing the sort operation). Reducing disk I/O is one primary focus of Oracle8 (or of any computer system for that matter), because disks operations are comparatively slower; they involve movement of mechanical components and should be optimized to the fullest possible extent. Oracle8 performs the sort operation in memory, allocating memory up to the maximum specified by the initialization parameter SORT_AREA_SIZE. If the volume of the data to be sorted is more than the SORT_AREA_SIZE, it uses temporary disk segments. The amount of disk space it uses depends mainly on the volume of the data to be sorted. When does Oracle perform sort operations? Oracle8 performs sorts while creating indexes and executing SQL operations that contain order by, group by, distinct, union, join, unique, and aggregate operations such as max, min, and so on. The analyze command also sorts data to calculate statistics.

Understanding Sort Behavior

The V$SYSSTAT dynamic performance table contains the statistics sorts (memory) and sorts (disk), indicating the number of in-memory and on-disk sorts the instance has performed since instance startup. It also contains the sorts (rows) statistics, which indicates the total number of rows sorted by the instance. Use the following query to see these statistics: SQL> column name format a20 SQL> select name,value from v$sysstat where name like 'sort%' NAME -------------------sorts (memory) sorts (disk) sorts (rows) Calculating the in-memory sort ratio In an OLTP environment, the sorts (disk) number should be a very low percentage of the total sorts. In Memory Sort Ratio can be calculated as In Memory Sort Ratio = 'sorts (memory')/( 'sorts (memory)' + 'sorts (disk)'). A value very near 100 percent is recommended for In Memory Sort Ratio. A lower value of this ratio indicates that the workload is performing too many disk sorts. In this system there were 11,462 sorts in the memory. Three sorts were done by using the disk. VALUE ---------11462 3 5187983

Optimizing Sort Operations

Sorts are one of the most resource-intensive operations in any database. If sorts are taking too much time on your system and are a cause of concern, consider the following options to minimize their impact: q Eliminate the sorting operation needed during the index creation by using the NOSORT clause of the CREATE INDEX command. To use NOSORT, you must presort the data before loading it into the database. Operating-system sort utilities as well as third-party sorting tools such as SYNCSORT can be used to sort the data before loadi ing it into the database. q If you're recreating an existing index, use the ALTER INDEX...REBUILD command to recreate the index by using the existing index rather than the table data: SQL> alter index I_orders rebuild ; Use the ANALYZE command with the ESTIMATE clause rather than with the COMPUTE clause:

http://www.informit.com/content/0789716534/element_018.shtml (10 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8

SQL> analyze table orders estimate statistics ; Ensure that the application isn't performing unnecessary sorts by using clauses such as distinct and union in the SQL commands when not needed. Several third-party tools do so very frequently. Trace the suspected queries and look at their execution plan to see if they include a sort operation.

Setting Sort-Related Parameters

It's certain that the workload will need some amount of sorting. You can minimize the disk sorts or enhance performance of disk sorts by using the following Oracle8 parameters to appropriate value depending on the application nature: q SORT_AREA_SIZE indicates the number of maximum memory in bytes a user process can allocate for sorting. Sort memory is allocated on a per-session basis. Each parallel query slave performing sorts has its own sort memory, thus the amount of memory used during parallel execution may increase drastically if this parameter is set too high. Ensure that the system has sufficient memory to meet the increased demand before increasing this parameter. q SORT_AREA_RETAINED_SIZE determines the amount of memory retained by the process after completion of a sort. The additional memory used for a sort is released back to the session. (Remember, it's not freed to the operating system.) This parameter has more significance when a session can perform multiple concurrent sorts, especially while using TP monitors or MTS. The default value of 0 for this parameter indicates that it's equal to the parameter SORT_AREA_SIZE. q Setting SORT_DIRECT_WRITES to TRUE enables the session to allocate additional buffers in memory and to write them directly to disk. This eliminates the use of a buffer cache for sorting operations and increases the performance. When this parameter is set to its default value of AUTO, Oracle8 enables direct sort writes when the SORT_AREA_SIZE is more than 10 times the database block size. Buffer memory used for SORT_DIRECT_WRITES is used from the session memory. q Set SORT_READ_FAC for the disk subsystem in the system to optimize the I/O performance during disk sorts. The formula to calculate SORT_READ_FAC is as follows: SORT_READ_FAC=(avg. seek time + avg. latency time + block transfer rate)/(block transfer time) Should you set SORT_DIRECT_WRITE to TRUE or AUTO? The values of SORT_WRITE_ BUFFER_SIZE and SORT_WRITE_BUFFER are considered only when SORT_DIRECT_WRITE is set to TRUE. The default values of these parameters are effective if SORT_ DIRECT_WRITE is set to AUTO.
q q

SORT_WRITE_BUFFER_SIZE determines the size of the I/O buffer used to perform direct sort writes. SORT_WRITE_BUFFER determines the total number of buffers used for direct sort writes.

Managing Temporary Segments

Oracle8 allocates temporary segments used during sorts from the tablespace defined as the user's temporary tablespace. Because the default value for this tablespace is SYSTEM, you must change it to some other tablespace. Improving sort performance To improve the performance of sort operations, assign a TEMPORARY-type tablespace as the user's temporary tablespace, setting the DIRECT_SORT_WRITE initialization parameter to TRUE, or setting the value of initialization parameter SORT_AREA_SIZE to a larger value. A production system will often need one or several tablespaces dedicated as temporary. To improve performance, you should alter these tablespaces to be of type TEMPORARY. If you use a permanent tablespace as temporary tablespace for sorting, each user has his dedicated sort segments deallocated at the end of the current operation. If you use a TEMPORARY-type tablespace as the temporary tablespace for users, the segments, once allocated by the instance, aren't deallocated and can be reused by other users who have been assigned to the same tablespace. This saves the overhead of allocating and deallocating the extents during the sort operations. It's recommended that you stripe the temporary tablespace to improve I/O performance during the sort write operations. Striping the temporary tablespace on the number of disks equal to the SORT_DIRECT_BUFFERS will yield the maximum performance benefit.

http://www.informit.com/content/0789716534/element_018.shtml (11 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Using Indexes, Clusters, Caching, and Sorting Effectively From: Using Oracle8

View disk sort usage statistics The dynamic performance view V$SORT_SEGMENT provides the usage statistics for the TEMPORARY tablespace. The V$SORT_USAGE view contains statistics related to the current sort activity in the instance. Set the initial and next extent parameters for the temporary tablespace as multiples of the SORT_AREA_SIZE in order to optimize extent allocation. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_018.shtml (12 of 12) [26.05.2000 17:09:38]

informit.com -- Your Brain is Hungry. InformIT - Improving Throu...h SQL, PL/SQL, and Precompilers Raman Batra From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Improving Throughput with SQL, PL/SQL, and Precompilers Raman Batra

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

Understanding SQL Tuning

r r r r r

Defining System Performance Deciding Which SQL Statements to Tune SQL Statement Tuning Index Tuning Data Tuning Keeping Shared SQL in the Shared Pool Speeding Up Access to the Shared SQL Area How to Identify Unnecessary Parse Calls Reducing Unnecessary Parsing Using Array Processing

Shared SQL
r r

SQL and PL/SQL Parsing Reduction

r r r

Understanding SQL Tuning

Because SQL is the most widely used utility for database queries, having a good handle on how to write efficient queries can give your system a great performance boost. You need to determine which database access method will yield the fastest system performance and use the least amount of system resources. The true power of SQL How SQL queries are written have the most impact on the performance of the database. Many times, a small change to a SQL query can make it execute many times more quickly-or slowly. Just because a SQL script accomplishes the task it was assigned to do doesn't in any way mean that it's a good script. A good SQL script is one that uses all the database features available to accomplish the task as fast as possible with as little system resource use as possible. That's where SQL tuning comes into play-how you can increase system throughput by reducing resource usage and contention.

Defining System Performance

Before discussing how to maximize system performance through tuning, you need to answer the question of

http://www.informit.com/content/0789716534/element_019.shtml (1 of 10) [26.05.2000 17:10:26]

informit.com -- Your Brain is Hungry. InformIT - Improving Throu...h SQL, PL/SQL, and Precompilers Raman Batra From: Using Oracle8

what exactly good system performance is. The easiest answer is to get the most amount of processing done in the least amount of elapsed time, but that's not always true. Good system performance is a balance between effective, efficient application code versus available system resources. By using threads and multiple CPUs, a single database application can use all the resources of an entire system. This results in the application executing very quickly, but other users of the system will most probably experience slow or sluggish response time. You create good system performance by writing efficient code and also by considering how system resources are affected. When thinking about system performance as it relates to software development with Oracle, what you need to keep in mind is how you structure your queries. It's true that database design and maintenance pretty much determines how a database will perform, but you can use a few tricks as a database administrator that will help your scripts and application programs run faster.

Deciding Which SQL Statements to Tune

Before you can begin tuning SQL statements, you need to decide which statements to tune. Your best bet is to start by using the utilities that come with the Oracle Server: TKPROF, Oracle Trace, and the EXPLAIN PLAN utility. You should also query the V$SORT_USAGE and V$SQLAREA views for more information on which SQL statements are using the most system resources. As a general rule, you should tune statements that do one of two things: take a very long time to execute or execute fairly quickly but are repeated many times (for example, in a loop). Tuning statements such as these will reap the most benefit in terms of system resources because they're using the most resources in your application. When tuning these statements, you want to reduce the resources the statements require, and, if possible, also reduce the number of times the statement is executed.

SQL Statement Tuning

This section focuses on quick suggestions on how to write the most effective SQL code and to perform the same task your code was doing before more quickly and efficiently. Use joins effectively How you order your joins in a SQL statement can affect the performance of the statement. When your SQL statement joins two or more tables, you'll want to list the tables that return the most rows first and have the smaller row selections later. You'll also want to use an index if possible to avoid full table scans. Take a Different Approach As stated earlier, SQL is a powerful language, and queries that produce the same result can be written in different ways. You can rewrite a query and then run the EXPLAIN PLAN command against the old and the rewritten query to determine the cost of each approach. In a nutshell, the utility analyzes a SQL statement and places the results in a table, which you query. The default name for this table is PLAN_TABLE, and its fields are as follows: SQL> describe plan_table Name Null? Type ------------------------------- -------- ---STATEMENT_ID VARCHAR2(30) TIMESTAMP DATE REMARKS VARCHAR2(80) OPERATION VARCHAR2(30) OPTIONS VARCHAR2(30) OBJECT_NODE VARCHAR2(128) OBJECT_OWNER VARCHAR2(30) OBJECT_NAME VARCHAR2(30) OBJECT_INSTANCE NUMBER(38) OBJECT_TYPE VARCHAR2(30) OPTIMIZER VARCHAR2(255) SEARCH_COLUMNS NUMBER ID NUMBER(38) PARENT_ID NUMBER(38) POSITION NUMBER(38)

http://www.informit.com/content/0789716534/element_019.shtml (2 of 10) [26.05.2000 17:10:26]

informit.com -- Your Brain is Hungry. InformIT - Improving Throu...h SQL, PL/SQL, and Precompilers Raman Batra From: Using Oracle8

COST CARDINALITY BYTES OTHER_TAG PARTITION_START PARTITION_STOP PARTITION_ID OTHER SEE ALSO How to invoke the EXPLAIN PLAN utility, Avoid Transformation Where Possible

NUMBER(38) NUMBER(38) NUMBER(38) VARCHAR2(255) VARCHAR2(255) VARCHAR2(255) NUMBER(38) LONG

If you can, avoid using the SQL functions that cast column information to a different data type-for example, numerical text information to a NUMBER. Also, avoid character information where you pull out a specific range of characters. When using the WHERE clause, make sure that the comparison is done on two variables of the same data type. A good example of what to avoid is the case where you're comparing an integer value with a character numerical value. By default, the optimizer will convert the character data to an integer before comparing, which will cause the SQL statement to fail if a row of character data doesn't translate to an integer. To avoid this, you should cast the integer value as a character value by using the to_char function, as in the following WHERE clause: WHERE

Declared as a VARCHAR2 Declared as a NUMBER Keep SQL Statements Single Purpose A long time ago there was a book out that contained a collection of programs written in a single line of the BASIC programming language. Because BASIC can do multiple things within a single line, it was possible to encapsulate an entire program in the context of one programming line. Although these programs worked, they were probably some of the worst examples of software design that have ever existed. Remember that when writing your SQL queries. Although it's possible to have a single SQL statement perform multiple operations, it's not always the best thing when you're working on improving system performance. Keep your queries to the point and single-tracked; watch out for SQL statements that do different things based on the data being selected. These types of operations bypass the optimizer and in many cases don't run as efficiently as they should. Make Use of Optimizer Hints By using optimizer hints (such as FULL, ROWID, and HASH), you can steer the optimizer down a particular data access path. Because you know the data and application better than the optimizer does, sometimes you'll override what the optimizer wants to do. By using hints, you can change the way the optimizer accesses data and therefore possibly speed up a query. (You can also slow down a query if it's not done right, so be careful!) Create Extra Tables for Reference When Necessary Sometimes you need to use the IN clause to select a group of rows. Suppose you're working at the Internal Revenue Service, and you want to list all the athletes you have in your database. Your query would look something like the following: Select name, gross, occupation from taxpayer where occupation in ('BOWLER', 'GOLFER', 'BASEBALL PLAYER', 'FOOTBALL PLAYER');

http://www.informit.com/content/0789716534/element_019.shtml (3 of 10) [26.05.2000 17:10:26]

informit.com -- Your Brain is Hungry. InformIT - Improving Throu...h SQL, PL/SQL, and Precompilers Raman Batra From: Using Oracle8

Do you see the problem? You would have a very long list of sports-related occupations to add, and the query would be very long and inefficient. Although there's really no way to tune the query, what you could do is create another table in the database called OCCUPATION_TYPE that would list each occupation name and the type of job that it is. The WHERE clause in your query now becomes WHERE Occupation_type = 'ATHLETE'; Be careful when using SELECT with IN Another possible performance killer is the use of a SELECT statement within an IN clause. You want to avoid the situation where SELECT could return hundreds or even thousands of rows, each one of which is processed with the IN clause. Not only will this run faster, but you won't have to worry about changing your query when a new type of sport gets added. Combine INSERT and UPDATE operations This suggestion is very simple to implement and has a good performance gain. When selecting data and updating it, you can combine those operations into a single SQL statement, so that the data being updated is read from the database only once. This reduces the number of calls to the database server, thereby increasing the performance of the query.

Index Tuning
To make the most out of database indexes, you might be adding more indexes or even dropping some. The common thought among software developers is that things will always run faster when using an index. But that's true only most of the time. If there are many indexes on a single set of Oracle data, there comes a point of diminishing returns, and the indexes could actually cause a performance bottleneck. So keep in mind that indexes aren't an answer to poorly written queries-they should be used with efficient code to get the most out of the Oracle Server. Consider hash clusters when indexing Hash clusters group data in the tables by applying a hash algorithm to the cluster key value of each table row. Hash clusters are very useful in the situation where tables are accessed with a WHERE clause. Performance can be better than just using an index.

Data Tuning
Believe it or not, how you store the raw data within the database can affect performance. This is particularly true when you're dealing with distributed data and are performing joins over the network. If that's the case, you can minimize processing time by keeping the data local as well by using replication. Also, partitioning the data helps as well by distributing the data across multiple disk drives that can be read in parallel.

Shared SQL
Sometimes you have different users executing the same or similar queries against the database. Each statement is parsed out and uses its own space in the shared pool. If two users submit the exact same query, however, Oracle stores only one copy of the parsed query in the SGA, and that one copy is shared among the processes making that query. This is a performance gain because of the space that this frees up in the SGA-space that can be used for other processing. Put common SQL into the shared pool Take your most common SQL and create stored procedures, which are automatically shared when multiple users execute them. Stored procedures also are parsed only once and stored in memory as parsed. Performance is improved through the elimination of reparsing. Look at the steps that Oracle takes to determine if a query already has an identical copy resident in the shared area: q First, Oracle takes the query and compares its hash value to that of all the other statements in the shared pool. If it finds a match, Oracle then compares the text of the query to the text of the other queries in the shared pool. The following group of SQL statements all return the same data, but are handled as

http://www.informit.com/content/0789716534/element_019.shtml (4 of 10) [26.05.2000 17:10:26]

informit.com -- Your Brain is Hungry. InformIT - Improving Throu...h SQL, PL/SQL, and Precompilers Raman Batra From: Using Oracle8

completely different by Oracle and don't share space in the SGA: SQL> SQL> SQL> SQL> SELECT SELECT select Select PLAYER, PLAYER, player, player, POSITION FROM TEAM; POSITION FROM TEAM; position from team; position from team;

Keep SQL text consistent When Oracle compares the SQL text, to find a match with an existing statement it must match exactly. The case of each character must be the same, and spaces must match exactly as well. If not, Oracle treats the text as two distinct SQL statements.
q

Oracle then compares the objects referenced in the statement to make sure that they're referencing the same physical objects within the database. Suppose that two users have loaded the demo tables into their personal schemas and are performing the same query on the EMP table. Even though the text is exactly the same, the query is referencing two distinct tables and is treated as two separate queries by the database. Lastly, the bind variable names in each statement-if they're used-must match exactly. The following shows two SQL statements that do the same thing but aren't shared in the pool: SQL> SELECT * FROM TEAM WHERE NUMBER = :PLAYERNUM; SQL> SELECT * FROM TEAM WHERE NUMBER = :PLAYERNO; Although the bind variable in the example references the NUMBER field of the TEAM table, it won't be shared in the shared pool because the two queries aren't syntactically the same. Develop a common naming convention for bind variables.

The benefit of programming standards If you get your developers together and decide on some common approaches to writing programs that access the Oracle Server, you can take advantage of being able to share SQL code within the SGA. Although it might be beneficial to change applications so that they use the same queries, you could define a few standards the developers would use and if they're submitting the same query. For example, you can define whether to use upper-case or lowercase and how they're spaced. Oracle will detect the similarity and share them because they will match all the criteria for sharing SQL. The benefit of shared SQL isn't really that two statements are sharing space in the shared pool; the real benefit comes when you have many application users running the same queries. That's why it's important to keep in mind shared SQL when writing your applications. When you get into the hundreds of users and the same SQL statements aren't shared among the users, you'll end up increasing the size of your shared pool unnecessarily. Sizing the shared pool Having a large shared pool takes away some of the issues regarding fitting objects into the shared pool. The INIT.ORA parameter SHARED_POOL_SIZE sets the size of the pool. You should make it as large as is reasonable, without making it so big that you waste system memory.

Keeping Shared SQL in the Shared Pool

Now that you've created identical SQL statements and they're being shared, you'll want to take steps to help ensure that they stay in the shared pool. You want to make sure that they're not knocked out of the shared pool because of the LRU (least recently used) algorithm used on the shared pool to keep the pool current. You also want to make sure that a large PL/SQL package being read into the shared pool doesn't overwrite your shared SQL statements. Take a quick look at how to prevent these two things from happening. How to Handle Large Allocations of Shared Pool The problem here is when a large PL/SQL script is being loaded for the first time, or that same script was loaded, was knocked out of shared pool because of non-use, and is being reloaded. This has the tendency to cause some of the smaller SQL statements in the shared pool to be knocked out of the shared pool because of the LRU algorithm. What you can do to prevent this is to use the INIT.ORA parameters to reserve space for large allocations. If this is done, the large allocations will be done in a reserved section of the shared pool and won't overwrite the other

http://www.informit.com/content/0789716534/element_019.shtml (5 of 10) [26.05.2000 17:10:26]

informit.com -- Your Brain is Hungry. InformIT - Improving Throu...h SQL, PL/SQL, and Precompilers Raman Batra From: Using Oracle8

statements in the pool. This is done with two initialization parameters: SHARED_POOL_RESERVED_SIZE and SHARED_POOL_RESERVED_SIZE_MIN_ALLOC First, set SHARED_POOL_RESERVED_SIZE to reserve a chunk of shared pool memory for large allocations. Next, set SHARED_POOL_ RESERVED_SIZE_MIN_ALLOC to the smallest value you want allocated in the reserved space. This means that any request for shared pool larger than the SHARED_POOL_ RESERVED_SIZE_MIN_ALLOC parameter will be allocated in the reserved memory, thus protecting the smaller statements from being pushed out. Keeping Objects in the Shared Pool Even if you use the parameters listed in the previous section, this won't prevent commonly used SQL statements from being aged out of the shared pool. There's something you can do about it, however. You'll want to check out the package DBMS_SHARED_POOL package. You create this package and package body by executing the DBMSPOOL.SQL and PRVTPOOL.PLB scripts, located in the /rdbms/admin directory under ORACLE_HOME. When you use this package, you can load objects into memory early, before memory fragmentation begins, and they can stay there for the duration. It's the DBMS_SHARED_POOL package that you use to pin a SQL or PL/SQL area. By pinning (locking in memory) large objects, you increase system performance in two ways: q Response time doesn't slow as the larger objects are read into the shared pool. q It's less likely that the smaller SQL areas will be aged out to make room for a much larger one. To pin objects in the shared pool, decide what you want to pin, start up the database, and then run DBMS_SHARED_POOL.KEEP. Three procedures come with the DBMS_SHARED_POOL package: q DBMS_SHARED_POOL.SIZES displays the objects in the shared pool that are larger than the size passed in as a parameter. q DBMS_SHARED_POOL.KEEP pins a SQL or PL/SQL area in memory. q DBMS_SHARED_POOL.UNKEEP marks a pinned object as available to be aged out. It's the opposite of the KEEP function, but doesn't actually remove the pinned object from the shared pool.

Speeding Up Access to the Shared SQL Area

Now that you know how to get your SQL into the shared pool and keep it in memory, look at how you can increase the speed of the shared SQL area of the SGA. An INIT.ORA parameter called CURSOR_SPACE_FOR_TIME can speed up execution calls on systems that have a large shared pool and plenty of free memory. It's a parameter set to TRUE or FALSE, and should be used only if your library cache misses on execution calls whose count is 0. If you have no library cache misses and you set the CURSOR_SPACE_FOR_TIME parameter to TRUE, you'll see a slight performance improvement on execution calls. The way it works is as follows: When the parameter is set to TRUE, a shared SQL area won't be deallocated until all cursors associated with the shared SQL statement have been closed. This saves some time because Oracle doesn't have to check to see if the statement is still in the shared pool when performing an execution call. Keep an eye on library cache misses You don't want to set CURSOR_SPACE_FOR_TIME to TRUE if you're seeing library cache misses. Performance will get worse if you do, and statements may begin failing with out-of-shared-pool errors.

SQL and PL/SQL Parsing Reduction

Another tool you can use to increase throughput of your SQL and PL/SQL objects is to tune the private SQL and PL/SQL areas. You can do this by reducing the number of parse calls made to the database. Again, this tuning technique is for systems with available free memory, because reducing parsing by the means suggested in the following sections will require more memory and a larger shared pool.

How to Identify Unnecessary Parse Calls

You need to find out the number of SQL execution calls being parsed compared with the number of statements being executed. Two Oracle tables can help you in this task: V$SQLAREA and V$SESSTAT.

http://www.informit.com/content/0789716534/element_019.shtml (6 of 10) [26.05.2000 17:10:26]

informit.com -- Your Brain is Hungry. InformIT - Improving Throu...h SQL, PL/SQL, and Precompilers Raman Batra From: Using Oracle8

V$SQLAREA The V$SQLAREA view contains information on the SQL statements in the shared SQL area. One row in this view constitutes one SQL statement. It provides statistics on SQL statements already in memory, parsed and ready for execution. The V$SQLAREA view contains the following field names: SQL> describe V$SQLAREA; Name Null? Type ------------------------------- -------- ---SQL_TEXT VARCHAR2(1000) SHARABLE_MEM NUMBER PERSISTENT_MEM NUMBER RUNTIME_MEM NUMBER SORTS NUMBER VERSION_COUNT NUMBER LOADED_VERSIONS NUMBER OPEN_VERSIONS NUMBER USERS_OPENING NUMBER EXECUTIONS NUMBER USERS_EXECUTING NUMBER LOADS NUMBER FIRST_LOAD_TIME VARCHAR2(19) INVALIDATIONS NUMBER PARSE_CALLS NUMBER DISK_READS NUMBER BUFFER_GETS NUMBER ROWS_PROCESSED NUMBER COMMAND_TYPE NUMBER OPTIMIZER_MODE VARCHAR2(25) PARSING_USER_ID NUMBER PARSING_SCHEMA_ID NUMBER KEPT_VERSIONS NUMBER ADDRESS RAW(4) HASH_VALUE NUMBER MODULE VARCHAR2(64) MODULE_HASH NUMBER ACTION VARCHAR2(64) ACTION_HASH NUMBER SERIALIZABLE_ABORTS NUMBER For your purposes, in this section you'll be concerned only with the SQL_TEXT, PARSE_CALLS, and EXECUTIONS fields. The SQL_TEXT field contains the actual text of the SQL statement being executed. The PARSE_CALLS field is the number of times the statement has been parsed, and the EXECUTIONS field is the number of times the statement was executed. Try the following SQL_TEXT query on one of your databases to get the parsing information: SQL> SELECT SQL_TEXT, PARSE_CALLS, EXECUTIONS 2> FROM V$SQLTEXT; Interpreting parses versus executions Look carefully at the ratio of parses to executions of the individual SQL statement. If the number or parses for any statement is close to the number of executions, that means you're continually reparsing that statement. V$SESSTAT The V$SESSTAT view stores statistics on the individual user sessions. The individual statistics are stored by number, so you need to query the V$STATNAME view with the statistic name to get the corresponding number. In Oracle8, this table contains more than 200 statistic names. For now, however, you'll be concerned only with the PARSE_COUNT and EXECUTE_COUNT statistics. First, get the statistic numbers for parse and execute counts. Start by executing the following two queries:

http://www.informit.com/content/0789716534/element_019.shtml (7 of 10) [26.05.2000 17:10:26]

informit.com -- Your Brain is Hungry. InformIT - Improving Throu...h SQL, PL/SQL, and Precompilers Raman Batra From: Using Oracle8

SQL> SELECT STATISTIC#, NAME FROM V$STATNAME WHERE NAME IN ('parse count (hard)','execute count'); You'll see output such as this: SQL> SELECT STATISTIC#, NAME FROM V$STATNAME 2> WHERE NAME IN ('parse count (hard)','execute count'); STATISTIC# ---------153 154 NAME -----------------------------------------------parse count (hard) execute count

From this output you can tell that the parse count statistic is number 153 and the execute count statistic is number 154. Now look at the V$SESSTAT view, which has only three fields: SQL> describe v$sesstat; Name Null? ------------------------------- -------SID STATISTIC# VALUE

Type ---NUMBER NUMBER NUMBER

Select the parse and execution statistics for all connected Oracle sessions with the following query: SQL> select * from v$sesstat 2 where statistic# in (153,154) 3 order by sid, statistic#; This query lists out (by session) the parse count and execution count for each session. You're looking for sessions where the two statistical values are closer rather than further apart. The closer in value the two statistics are, the more potential there is to reduce unnecessary parsing. The output of the query is as follows: SID STATISTIC# VALUE ---------- ---------- ---------1 153 0 1 154 0 2 153 0 2 154 0 3 153 0 3 154 0 4 153 0 4 154 0 5 153 0 5 154 0 6 153 5 SID STATISTIC# VALUE ---------- ---------- ---------6 154 46 7 153 1 7 154 14 8 153 52 8 154 395 16 rows selected. In the output, look at session ID 8; it had 395 execution calls but only 52 parses. You'll want to find sessions where the parse count is much higher to reduce unnecessary parsing. (This query was run against the test database, which unfortunately doesn't have much of a user load, but this is the query you use to get this information.)

http://www.informit.com/content/0789716534/element_019.shtml (8 of 10) [26.05.2000 17:10:26]

informit.com -- Your Brain is Hungry. InformIT - Improving Throu...h SQL, PL/SQL, and Precompilers Raman Batra From: Using Oracle8

Oracle Performance Pack A great tool for interpreting data-base statistics is Oracle's Performance Pack, which provides a graphical interface to database statistics such as cache rates, disk I/O, and SQL performance.

Reducing Unnecessary Parsing

The key to eliminating unnecessary parsing is how you handle the cursors. This is especially true when you're using explicit cursors. When using explicit cursors, be sure to avoid the situation where you open and close cursors in a loop. Moving the open and close statements outside the loop will keep the statement from being reparsed. Your program also should DECLARE the cursor, reOPEN it each time the value of a host variable changes, and CLOSE the cursor only after the SQL statement is no longer required. Remember also that when you're using MODE=ANSI with the precompilers, a COMMIT command will close a cursor. The cursor-management functions HOLD_CURSOR, RELEASE CURSOR, and MAXOPENCURSORS give you additional power in how to control the parsing and reparsing of SQL statements. In a nutshell, these commands help you control how Oracle handles the creation and uses cursors. You use these commands to keep cursors to prevent reparsing and to remove cursors to save memory.

Using Array Processing

Using host arrays in a program is useful when you're updating multiple rows of a table. You get a real performance gain if the rows are contiguous. You define the arrays the same way you would any other programming array. For example, look at an array of size 100 for the three fields of the table TEAM: int char int uniform_no[100]; player_name[100][25]; position[100];

Oracle8 supports only single-dimensional arrays. The player_name array is single dimensional; the 25 is simply the maximum size of each string. The benefit here comes in how you can update the table after the arrays are populated. Rather than use a FOR loop to update the table in the following example: FOR (counter = 0; counter <=100; counter++) EXEC SQL INSERT INTO TEAM (uniform_no,player_name,position) VALUES(:uniform_no[counter], :player_name[counter], :position[counter]); you could pass just the array name to the EXEC statement. Oracle not only updates the table with all the elements in the arrays, but also does it as a single operation. Handling NULL values in a program You also can use indicator arrays to assign NULLs to variables in input host arrays, and to detect NULL or truncated values in output host arrays. Check the Oracle documentation for more information on indicator arrays. In the preceding example, 100 insert statements are executed. Look at the new code: EXEC SQL INSERT INTO TEAM (uniform_no,player_name,position) VALUES(:uniform_no, :player_name, :position); The loop is no longer necessary, and you have a much nicer looking piece of code.

http://www.informit.com/content/0789716534/element_019.shtml (9 of 10) [26.05.2000 17:10:26]

informit.com -- Your Brain is Hungry. InformIT - Improving Throu...h SQL, PL/SQL, and Precompilers Raman Batra From: Using Oracle8

< Back

Contents

Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_019.shtml (10 of 10) [26.05.2000 17:10:26]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Tuning Your Memory Structures and File Access

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q

Why Oracle Must Be Tuned Database Buffer Cache

r r

Examining Performance Testing New Buffer Cache Settings Examining Library Cache Performance Examining Data Dictionary Cache Performance Setting New Shared Pool Parameter Values Examining Performance Comparing Dedicated Versus Shared Servers Managing Sort Space Locating Data File Hot Spots Using Striping Strategies Sizing Redo Logs to Allow Checkpoint Completion Sizing Redo to Allow Archive Completion Designing Redo to Avoid LGWR and ARCH Contention

Tuning the Shared Pool

r r r

Redo Log Buffer

r

Process Global Area

r r

Managing Data File Contention

r r

Maximizing Redo Log Performance

r r r

Why Oracle Must Be Tuned

One of Oracle8's strongest assets is its rich tuning facility. Oracle8 offers a staggering array of parameters that you can tune to achieve maximum performance for any database system. Because every database has different data structures and different transaction loads, you can maximize the return on your hardware investment only by tuning Oracle. This will minimize the total cost of ownership for your database system and provide users the maximum performance Oracle can deliver. Although there may be no need initially to dive into the full array of tuning parameters for acceptable performance, the need will become more acute as a database becomes larger and services more transactions.

http://www.informit.com/content/0789716534/element_020.shtml (1 of 14) [26.05.2000 17:11:33]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

Oracle, out of the box, is tuned to run acceptably well on systems with low memory and CPU resources; most databases of any significant size should be tuned before being considered production quality.

Database Buffer Cache

Disk I/O operations are hideously expensive when compared to I/O operations performed in memory. In fact, reading information from memory is often thousands of times faster than when reading from disk. Oracle capitalizes on this by caching the most recently used database blocks in a memory segment that resides in the SGA. This memory block is known as the database buffer cache, or sometimes as just the buffer cache. Beware of swapping When allocating a larger database buffer cache, pay attention to whether the operating system is swapping real memory pages to disk to alleviate a memory shortage. Swapping is terribly expensive and will, by far, outweigh any benefit you gain from an oversized buffer cache. When tuning the database buffer cache, bigger is usually better. The more database blocks that Oracle can keep in memory, the faster it will perform. When allocating more memory to the database buffer cache, be aware that as the buffer hit ratio approaches 100 percent, it will take increasingly larger amounts of additional memory for the buffer cache's hit ratio to increase noticeably. Most active database systems have a discernible working set of very frequently used database blocks. When tuning a database, the goal is to keep at least the working set of database buffers in memory to maintain acceptable performance. Not surprisingly, as a database grows larger, it will need a larger buffer cache to maintain the same hit ratio.

Examining Performance
We're interested in knowing the hit ratio of the buffer cache for a database. To determine this, we must know the number of q Consistent gets q Logical block reads q Physical block reads Consistent gets Consistent gets are logical block reads associated with Oracle's read consistency system. When rollback segment blocks must be read to provide read consistency, they're not included in the normal logical block reads statistic. The hit ratio is computed as follows: (consistent gets + logical reads - physical reads) / (consistent gets + logical reads) The V$SYSSTAT table contains all the needed information since the database was last started. This table is easily accessed through SQL*Plus; the following example uses SQL*Plus to find the database buffer cache hit ratio:

http://www.informit.com/content/0789716534/element_020.shtml (2 of 14) [26.05.2000 17:11:33]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

FROM V$SYSSTAT; Hit Ratio ---------77.0755025 1 row selected. SQL> Used to read rows from V$SYSSTAT table Used to read the value associated with each specified parameter Ideally, a database running OLTP-type transactions should see a hit ratio of 90 percent or more. During long-running batch jobs, the hit ratio may fall into the 70-80 percent range. Anything lower than 70 percent may indicate that a larger database buffer cache is needed. Many DBAs who are familiar with other database systems on the market will find queries against V$ views to gather performance statistics crude and time-consuming. Fortunately, Oracle has companion packs for its Enterprise Manager system that will show, graphically, a database's performance health. These tools can also go a long way toward helping you tune your database(s) for optimum performance. Third-party vendors, including BMC and Platinum, offer similar tools that also work quite well for non-Oracle databases; these tools are particularly well suited for DBAs supporting a heterogeneous database environment.

Testing New Buffer Cache Settings

Allocating a larger database buffer cache often also means adding memory to the database server. Before writing a purchase order, it's often useful-if not necessary-to know the benefit more memory will bring to the database. Impact of DB_BLOCK_LRU_STATISTICS Be aware that DB_BLOCK_LRU_EXTENDED_STATISTICS has a noticeable impact on Oracle's internal overhead. Don't run DB_BLOCK_LRU_EXTENDED_STATISTICS any longer than you must to gather the needed statistics. Oracle8 has a utility that lets you see the effect on the buffer cache hit ratio after a hypothetical change in the size of the buffer cache. By enabling the DB_BLOCK_LRU_EXTENDED_STATISTICS parameter in the INIT.ORA file, Oracle8 will record the hypothetical statistics in the following tables: q x$kcbrbh records "what-if" statistics to show the number of additional cache hits for each additional block added to the database buffer cache. In this table, the following fields are of interest: INDX COUNT
q

Identifies each new phantom block for which you're collecting statistics Indicates how many additional cache hits would occur if the block number INDX were added to the database buffer cache

x$kcbcbh records hypothetical statistics to show the effect of reducing the size of the database buffer cache. As this utility is very rarely used, its use isn't covered in this chapter. Oracle's documentation should be consulted for further information. Enable DB_BLOCK_LRU_EXTENDED_STATISTICS

http://www.informit.com/content/0789716534/element_020.shtml (3 of 14) [26.05.2000 17:11:33]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

1. Edit the appropriate INIT.ORA file and add the following line: DB_BLOCK_LRU_EXTENDED_STATISTICS = additional buffers 2. Shut down the database. 3. Restart the database. Disable DB_BLOCK_LRU_EXTENDED_STATISTICS 1. Edit the appropriate INIT.ORA file and remove the following line: DB_BLOCK_LRU_EXTENDED_STATISTICS = number 2. Shut down the database. 3. Restart the database. Calculating a New Hit Ratio from Additional Buffer Cache The database is now running an OLTP application that achieves only an 81 percent database buffer cache hit ratio. Let's test the effect on the hit ratio by adding 2,048 blocks to the buffer cache. First, you set the number of hypothetical new buffers. Use an operating system editor to edit the INIT.ORA file to include the following line: DB_BLOCK_LRU_EXTENDED_STATISTICS = 2048 Next, shut down and restart the database. Server Manager is run from the operating system command line. The shutdown and startup session is as follows: SVRMGR> connect internal; Connected. SVRMGR> shutdown immediate; Database closed. Database dismounted. ORACLE instance shut down. SVRMGR> startup ORACLE instance started. Database mounted. Database opened. SVRMGR> At this time, either normal OLTP database operations should resume or test scripts should be run to simulate normal database load. In either case, make sure that enough realistic activity occurs for the database buffer cache hit ratio to stabilize. Now calculate the number of additional hypothetical cache hits for all 2,048 phantom database buffers, as shown in this Server Manager session: SVRMGR> SELECT SUM(COUNT) FROM X$KCBRBH; SUM(COUNT) ---------43712 1 row selected. SVRMGR> Using X$ tables X$ tables are owned by the SYS user and usually don't have public synonyms established. These examples use Server Manager because connect internal always connects as the SYS user. You can run these same queries from SQL*Plus, as long as you log in as the SYS user.

http://www.informit.com/content/0789716534/element_020.shtml (4 of 14) [26.05.2000 17:11:33]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

We now know that 43,712 additional cache hits would occur if the buffer cache were enlarged by 2,048 blocks (rather than force physical reads). Finally, you need to calculate the hypothetical cache hit ratio. Let's use the same formula used in the last section to compute the cache hit ratio, but this time subtract 43,712 phantom cache hits from the actual physical reads as shown: SVRMGR> SELECT ((SUM(DECODE(NAME, 'consistent gets', 2> VALUE, 0))+ 3> SUM(DECODE(NAME,'db block gets',VALUE, 0)) 4> (SUM(DECODE(NAME,'physical reads',VALUE,0))) 5> 43712) / (SUM(DECODE(NAME, 'consistent gets', 6> VALUE, 0))+SUM(DECODE(NAME,'db block gets',VALUE, 7> 0))) * 100) "Hit Ratio" FROM V$SYSSTAT; Hit Ratio ---------92.2935720 1 row selected. SVRMGR> You can see here that adding 2,048 blocks to the database buffer cache would cause the cache hit ratio to increase from 81-92 percent. What's the next step? Knowing that there would be a very noticeable rise in the database buffer cache hit ratio, you would want to add 2,048 to the DB_BLOCK_BUFFERS parameter in the INIT.ORA file. Don't forget to turn off DB_BLOCK_LRU_EXTENDED_STATISTICS! Change the size of the database buffer cache 1. Use any standard operating system text file editor to change the DB_BLOCK_BUFFERS parameter in the INIT.ORA file to reflect the size of the buffer cache in database blocks. 2. Stop the database. 3. Restart the database. Example of Enlarging the Buffer Cache by 2,048 Blocks The current INIT.ORA file has the following parameter line: DB_BLOCK_BUFFERS = 8192 In the example from the preceding section, it was determined that adding 2,048 blocks would help improve database response time. Now edit the INIT.ORA file to contain these lines: Comment INIT.ORA file changes When you're diagnosing sudden changes in the database's performance or reliability, it's often invaluable to know what crucial parameters were changed recently. Any time parameters are changed in INIT.ORA, you should add at least a one-line description of the change, who made it, and when it was done. # # This setting was changed on 2 FEB 98 by TMG x3707. Buffer # Cache hit ratio rose from 81% to 92% by increasing it # 2048 blocks # DB_BLOCK_BUFFERS = 8192 # DB_BLOCK_BUFFERS = 10240 Whenever the DB_BLOCK_BUFFERS parameter is changed, the database will have to be shut down and restarted for the changes to take effect.

http://www.informit.com/content/0789716534/element_020.shtml (5 of 14) [26.05.2000 17:11:33]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

Tuning the Shared Pool

Oracle8's shared pool contains the following: q Execution plans for recent SQL statements, known as the library cache q Data dictionary cache q Some miscellaneous session information Recall that every SQL statement executed by the database engine must be parsed, evaluated, and executed. This process is relatively time-consuming, because Oracle8 has very sophisticated methods for determining the optimum execution plan. By caching execution plans and the data dictionary, Oracle8 can quickly execute similar SQL statements by retrieving preprocessed execution plans and data dictionary information from the shared pool. The goal of tuning the shared pool will be to achieve a +95 percent hit ratio on the library cache and a 95-100 percent hit ratio on the data dictionary cache during all operations.

Examining Library Cache Performance

The V$LIBRARYCACHE view contains information on the library cache, including cache hit statistics. The cache hit ratio is determined by issuing the following SQL statement: SELECT SUM(PINS)/(SUM(PINS)+SUM(RELOADS))*100 "Cache Hit Ratio" FROM V$LIBRARYCACHE; SQL*Plus will reply with a single row and column that will be the library cache hit ratio. Example of Calculating the Library Cache Hit Ratio The following command is issued after you log on to SQL*Plus as a DBA user (the SYSTEM user in this case): SQL> SELECT SUM(PINS)/(SUM(PINS)+SUM(RELOADS))*100 2 "Cache Hit Ratio" 3 FROM V$LIBRARYCACHE; Cache Hit Ratio --------------99.8039216 SQL> Here we can see the library cache hit ratio is almost 100 percent, which is ideal. If the library cache hit ratio were lower than 95 percent, you would probably want to add memory to the shared pool.

Examining Data Dictionary Cache Performance

Data dictionary cache statistics are kept in the V$ROWCACHE view, which categorizes data dictionary cache statistics. However, most DBAs will benefit most from knowing an aggregate hit/miss percentage with the following SQL statement: SELECT (1-(SUM(GETMISSES)/SUM(COUNT)))*100 "Dictionary Cache Hit Ratio" FROM V$ROWCACHE; This statement will return the data dictionary cache hit ratio since the database was last started. Example of Calculating the Data Dictionary Cache Hit Ratio The following command is issued after you log on to SQL*Plus as a DBA user (the SYSTEM user in this case): SQL> SELECT (1-(SUM(GETMISSES)/SUM(COUNT)))*100 2 "Dictionary Cache Hit Ratio" 3 FROM V$ROWCACHE;

http://www.informit.com/content/0789716534/element_020.shtml (6 of 14) [26.05.2000 17:11:33]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

Dictionary Cache Hit Ratio -------------------------63.1687243 SQL> Considerations with low hit ratios When a low hit ratio is encountered, always consider how long the database has been up and running. If a database was just started or started a while ago with little or no subsequent application activity, it's very likely that Oracle simply hasn't had the opportunity to cache the information it will need. Oracle caches information only when it's needed; all information contained in the cache was preceded by at least one cache miss. Notice that the dictionary cache hit ratio is only 63 percent-far below the ideal ratio of 99 percent and the acceptable threshold of 95 percent. In this case, the shared spool size almost certainly must be increased.

Setting New Shared Pool Parameter Values

Although we examine the shared pool in library cache and data dictionary cache steps, adding memory to either is accomplished by adding memory (in bytes) to the shared pool with the SHARED_POOL_SIZE parameter in the applicable INIT.ORA file. The following line would allocate 3,500,000 bytes (~3.5MB) to an instance's shared pool. This is the default setting found in an INIT.ORA file: Library versus data dictionary cache allocation Oracle8 automatically proportions the shared pool into data dictionary and library cache segments. You can't adjust their respective sizes manually. If either cache hit ratio falls below par, you must increase the size of the shared pool. shared_pool_size = 3500000 Choosing new sizes for the shared pool is something more akin to art than science. Based on the library and data dictionary cache hit ratios, you may need to increase the shared pool perhaps just 25 percent, or to many times the default size. When choosing new shared pool sizes, be aware that it will usually take a much larger increase in memory to increase a cache hit ratio from 90 to 95 percent than from 70 to 80 percent. Change the size of the shared pool 1. Modify the SHARED_POOL_SIZE parameter in the appropriate INIT.ORA file. The following sets the shared pool to 16MB: # # Original setting increased to 16MB. Library cache hit # ratio was 61%. # TMG: 21 Jan 98 # shared_pool_size = 3500000 SHARED_POOL_SIZE = 16777216 2. Stop the database. 3. Restart the database.

Redo Log Buffer

The redo log subsystem is probably one of the most heavily accessed components of Oracle; every change to the database system must go through it. Spooling all changes directly to disk would impair transaction throughput, so Oracle8 buffers redo log information in the SGA to better utilize precious I/O resources. Considerations for SMP systems In addition to the redo log buffer size, SMP database DBAs must also consider the number of latches needed to prevent latch contention when accessing the redo log buffer area. Consult Oracle's documentation for information on the V$LATCH view. It's important to size the redo log buffer appropriately to avoid user database processes from having to wait on
http://www.informit.com/content/0789716534/element_020.shtml (7 of 14) [26.05.2000 17:11:33]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

space in the buffer to continue operations. During high update transaction rates, the log writer process may not be able to keep up with several user processes all submitting redo log information that must be written. The redo log buffer holds redo log information so that user processes can continue; as the log writer process catches up, it will purge information from the redo log buffer.

Examining Performance
When tuning the redo log buffer, focus on these statistics in the V$SYSSTAT view: q redo log space requests is the number of times user processes had to wait for space to become available in the redo log buffer. Ideally, this parameter will be zero; in practice, however, it's acceptable for a few user processes to wait on space in large systems. Size the redo log buffer to keep redo log allocation waits close to zero. q redo buffer allocation retries is the number of times the redo writer had to stop and wait for the log writer to finish writing dirty buffers. Again, this parameter should be zero; however, a low number is acceptable on larger systems. Use the following SQL command to gather redo log statistics: SELECT NAME,VALUE FROM V$SYSSTAT WHERE NAME IN ('redo buffer allocation retries','redo log space requests'); The result of this query will be two rows listing the value of each parameter. Example of Redo Log Performance Evaluation The preceding command is issued and the result is shown in the following sample SQL*Plus session for a DBA user (SYSTEM in this case): SQL> SELECT NAME,VALUE FROM V$SYSSTAT WHERE NAME IN 2 ('redo buffer allocation retries', 3 'redo log space requests'); NAME VALUE ------------------------------------------------ ---------redo buffer allocation retries 835 redo log space requests 508 SQL> Because both statistics are quite high in this case, it is advisable to add memory to the redo log buffers. Set new redo log buffer parameters 1. Change the LOG_BUFFER parameter in the appropriate INIT.ORA file to the new size of the redo log buffer, in bytes. Any text file editor may be used for this purpose. 2. Stop the database. 3. Restart the database. Guidelines for LOG_BUFFER values Oracle's default will depend on your platform and database version. Typically, however, the default is too small. 64KB to 128KB will usually work out well for smaller databases. 512KB to 1MB (or even more) may be needed on larger database systems. Example of Setting LOG_BUFFER in INIT.ORA The following INIT.ORA file section was changed in an operating system-supplied text editor to increase the size of the LOG_BUFFER parameter from 16KB to 256KB: # # 16k Proved to be too small. Increased the size to 256k # TMG: 03 Mar 98 # LOG_BUFFER = 16384 # LOG_BUFFER = 262144

http://www.informit.com/content/0789716534/element_020.shtml (8 of 14) [26.05.2000 17:11:33]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

Process Global Area

Whenever a connection is made to an Oracle database, a unique section of memory is allocated to store session-level information. The area of memory allocated for each session is known as the Process Global Area (PGA). Tuning the PGA is largely done to allow Oracle to scale well based on hardware constraints and the needs and use patterns of the predominate applications. As more and more users begin to utilize a particular database, it will become more important to ensure that Oracle is tuned to scale as well as possible with a given hardware configuration.

Comparing Dedicated Versus Shared Servers

Oracle8 offers two different options for managing each session connected to the database: q Dedicated Server will create a new server process for each connection to the database. This is Oracle's default behavior. q Multithreaded Server creates a finite number of server processes shared between all the individual connections to the database. Use multithreaded only when necessary Many DBAs mistakenly believe that Oracle's multithreaded server (MTS) will perform better than the dedicated model. This is usually the result of contemporary lingo that almost always associates "multithreaded" with faster. In reality, MTS will slow a database's operation overall. MTS can drastically increase the number of concurrent database sessions-that is the only advantage it offers. Don't use MTS unless scalability requirements dictate its use. You can achieve the best performance by using the dedicated server process. Because no sharing of the server processes exists, when a database session needs to access the database, a ready server process is guaranteed to service its needs. In environments where there could be many hundreds or even thousands of simultaneous connections to the database, Multi-threaded Server can offer better scalability. Short and infrequent queries are often made to the database server in OLTP environments. In the dedicated server model, each session (which may go for hours between database activity) will have its own server process waiting for SQL commands to execute. Each server uses a quantity of system memory and a slot in the operating system's process table, so there's a real limit to the number of the server processes that may be running at any given time. By using a fixed pool of server processes, you can continue to add concurrent users without consuming any more system memory. Multithreaded Server's downside exists precisely for the reason it offers increased scalability. Each server process can service only one database connection at a time (despite the deceiving "multithreaded" adjective); if more connections need servers than there are servers available, the additional connections must wait until a server process becomes available. Obviously, perceived performance can suffer noticeably when a client must wait for a server process to become available. Setting up MTS Installing and configuring Multi-threaded Server can be a rather complex task that's outside the scope of this book. Refer to Oracle's product documentation for installation and configuration instructions.

Managing Sort Space

Sorting will occur if an SQL statement includes ORDER BY or GROUP BY. It can, however, also occur when joining tables and various other Oracle internal operations. Although much can be done to improve sorting performance, the best way to increase performance is to eliminate sorts in the first place when possible. When Oracle must sort a set of data, it must use a temporary workspace to create the sorted list. Oracle can use two locations: q Memory q A segment in the user's defined TEMPORARY TABLESPACE on disk As you can imagine, sorting in memory is much faster than sorting on disk. The SORT_AREA_SIZE parameter in the INIT.ORA file describes, in bytes, the amount of space available to each server process for performing sorts in memory.
http://www.informit.com/content/0789716534/element_020.shtml (9 of 14) [26.05.2000 17:11:34]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

The V$SYSSTAT view keeps statistics on the number of sorts performed in memory and on disk. Use the following SQL statement to retrieve sorting statistics: SELECT NAME,VALUE FROM V$SYSSTAT WHERE NAME LIKE 'sort%'; When to read statistics As with all other statistics, be sure to allow Oracle and the predominate applications to run for a fair amount of time before acting on any statistics gathered. If Oracle and the associated application(s) haven't run long enough (a full business day is typical), the values retrieved may not truly reflect the workings of the database. OLTP applications should perform almost no sorts on disk (user response is paramount). DSS systems may see larger sorts running on disk but, if possible, these too should be run in memory for the best possible performance. Tune the database for sorting 1. Query the V$SYSSTAT view to determine the number of sorts performed in memory and on disk. 2. If fewer sorts on disk are desired, increase the SORT_AREA_SIZE parameter in the appropriate INIT.ORA file. 3. Stop and restart the database. 4. Allow the database to run at least a full business day for accurate sorting statistics to be gathered. 5. Repeat this procedure until the number of sorts performed on disk is acceptable. Example of Tuning the Sort Space The following SQL*Plus session shows the V$SYSSTAT view being checked by the SYSTEM user: SQL> SELECT NAME,VALUE FROM V$SYSSTAT WHERE 2 NAME LIKE 'sort%'; NAME VALUE ------------------------------ ---------sorts (memory) 2922 sorts (disk) 97 sorts (rows) 32693 SQL> Because the primary application is OLTP in nature, the number of sorts performed on disk (97) is a bit high. The following line appears in the relevant INIT.ORA file: sort_area_size = 65536 The first test to reduce disk sorts will increase the sort_area_size parameter to 128KB. The new section will be as follows: # # 97 sorts were being performed on disk. Parameter # increased to 128k # TMG: 11 Apr 98 # sort_area_size = 65536 sort_area_size = 131072 After the database was stopped and restarted (to activate the new parameter), one full business day elapsed and the V$SYSSTAT view was queried again. These are the results: NAME VALUE ------------------------------ ---------sorts (memory) 3661 sorts (disk) 3 sorts (rows) 34014

http://www.informit.com/content/0789716534/element_020.shtml (10 of 14) [26.05.2000 17:11:34]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

By increasing the sort_area_size parameter to 128KB, the number of disk sorts has been lowered to an acceptable level.

Managing Data File Contention

Databases often exhibit the characteristic that most data that must be accessed is physically close together. Even though a database may be many gigabytes in size, only a few dozen megabytes or so may be accessed over an hour. If this information is physically close, it's likely that only a handful of database data files are actively being used. Reducing hot spots by design When designing a database, try to identify which segments will receive the highest volume of data accesses. Heavily accessed segments ideally should be placed in their own tablespaces with composing data files distributed across different disk drives and controllers. Data files accessed at much higher rate than most others over a particular time period are known as hot spots. From a tuning perspective, hot spots should be minimized if at all possible. Most database servers are designed with data files distributed over many disk drives to maximize the potential for parallel I/O. Hot spots result in just a few disk drives at 100 percent utilization, whereas most others sit idle. When a disk drive (or controller) is fully utilized, any further requests have to wait in a queue to be serviced. Fortunately, Oracle keeps statistics on each data file's I/O rates to help you identify hot spots. The V$FILESTAT contains the information needed to identify and reduce data file hot spots.

Locating Data File Hot Spots

The V$FILESTAT view will be queried for physical reads and physical writes on each data file. By looking at the number of physical reads and writes in each of a database's data files, you can discern which data files are hot spots. Be aware that the values reported for physical reads and writes are counted from the time the database was last started. To identify hot spots during a particular period of time, it's necessary to compare the contents of the V$FILESTAT view at the beginning and end of the time period of interest. The following SQL statement shows the physical read and write values for each data file sorted by the total number of physical accesses: select name,phyrds,phywrts,(phyrds+phywrts) Total from v$filestat s, v$datafile d where s.file#=d.file# order by total desc; V$FILESTAT and V$DATAFILE views This SQL code uses the V$FILESTAT and V$DATAFILE views. This is necessary because the statistics in the V$FILESTAT view are identified by file number only. By inner joining the V$FILESTAT and V$DATAFILE views, file statistics can be mapped to filenames. By looking at data files at the top of the result from this SQL statement, you can identify which physical devices are being accessed the most. If any of the top data files are on the same physical disk drives or controllers, you should consider moving them to different drives or controllers. Example of Identifying Hot Spots The V$FILESTAT view is queried for hot spot information by the SYSTEM user in SQL*Plus. The resulting session is as follows: SQL> select name,phyrds,phywrts,(phyrds+phywrts) Total 2 from v$filestat s, v$datafile d 3 where s.file#=d.file# 4 order by total desc; NAME PHYRDS PHYWRTS TOTAL ----------------------------- -------- ---------- -----/oracle/sapdata4/stabi.data2 322805 6373 329178

http://www.informit.com/content/0789716534/element_020.shtml (11 of 14) [26.05.2000 17:11:34]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

/oracle/sapdata1/stabd.data1 ... /oracle/sapdata5/user1i.data1 /oracle/sapdata2/docud.data1 /oracle/sapdata2/loadd.data1 /oracle/sapdata3/loadi.data1 /oracle/sapdata2/protd.data1 <

57173 11 6 6 4 0

2871 0 3 1 0 0

60044 11 9 7 4 0

These two data files reside on the same physical device (mounted on /oracle/ sapdata2) Because btabd.data1 and protd.data2 would be considered hot spots and both reside on the same physical device, it's possible they should be moved to different physical devices. As with all statistics, confirm that they are a representative sample before using them to guide any changes to the database.

Using Striping Strategies

Data-access patterns often unavoidably produce data file hot spots no matter how well the database is designed. You can use two key strategies to minimize data file contention caused by these hot spots: q Rebuilding tablespaces with more data files that are smaller in size and located on different disk drives and controllers q RAID technology The simplest and (often) cheapest remedy for contention is to recreate a tablespace identified as housing hot spots with more data files that are smaller in size. The key here is to keep data files smaller than they were in the original; this will force Oracle to spread extents out over more data files (and, therefore, physical devices) and minimize data file contention. This solution is most ideal on servers that have many individual disk drives not being heavily used. In recent years, RAID technology has been rapidly adopted in many database servers to maximize I/O throughput with little administrative work. Because the hardware automatically stripes physical (not database) storage blocks for each data file across many disk drives, no changes to the database are necessary. RAID technology can help reduce data file contention without causing you much (if any) additional workload. RAID technology isn't a magic cure Inexperienced DBAs sometimes mistakenly believe that RAID technology eliminates the need for them to properly design a data-base's physical storage layout (such as isolating tables and indexes on different tablespaces and sizing segments). Unfortunately, although RAID can offer perhaps two or even five times the throughput of a single disk, proper database design can offer overall performance increases of several magnitudes. RAID technology will offer the most benefit to properly designed databases. By implementing one or both of these strategies, the performance degradation caused by data file contention can be reduced to an acceptable level.

Maximizing Redo Log Performance

Oracle must keep a log of all its write activity in order to facilitate lossless recovery from instance and media failure. The redo logs contain exceedingly detailed information about most every change Oracle makes to a database's physical file components. Because so much information is written to the active redo log, it can easily become a bottleneck that can significantly limit the performance of an otherwise well-tuned database. When you're tuning the redo logs for performance, your objective will be to minimize the amount of time the database must spend writing redo information to disk.

Sizing Redo Logs to Allow Checkpoint Completion

Checkpoints, while necessary for system integrity, also can drastically affect system performance depending on their frequency and duration. The system should ideally be tuned such that a checkpoint occurs only during a log switch.
http://www.informit.com/content/0789716534/element_020.shtml (12 of 14) [26.05.2000 17:11:34]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8

Use the checkpoint process Most sites should use the optional checkpoint process to take care of updating data file headers during checkpoints. By using the checkpoint process, the log writer process won't be interrupted from writing redo log information during a check-point, thereby further reducing performance problems caused by checkpoint activities. When sizing redo logs, it's absolutely essential that they be sufficiently sized so that checkpoints can finish well before a log switch is necessary. Oracle's V$SYSSTAT view contains the statistical values for background checkpoints started and background checkpoints completed. When the values for these statistics vary by more than 1, at least one checkpoint didn't finish before a log switch was necessary and log sizes must be increased. Use the following SQL statement to query the V$SYSSTAT view for the background checkpoint statistics: select name,value from v$sysstat where name like 'background checkpoint%'; Two rows will be returned: one with the count of background statistics started and another with the count of background statistics finished. Example of Checking for Checkpoint Completion The V$SYSSTAT view is queried for the background checkpoint statistics by the SYSTEM user in SQL*Plus. The example session follows: SQL> select name,value from v$sysstat 2 where name like 'background checkpoint%'; NAME VALUE -------------------------------------------------- -------background checkpoints started 4521 background checkpoints completed 4520 SQL> Identifying the problem The two background checkpoint statistics vary by only 1. This indicates that checkpoints are finishing before a log switch is forced. If the background checkpoints completed were 3,788, we would have to increase the size of the redo log files until the two statistics varied by only 1.

Sizing Redo to Allow Archive Completion

When a database is running in ARCHIVELOG mode and a log file switch occurs, the database will immediately begin filling a new redo log group as the ARCH process begins copying one of the completed redo logs to the archive log destination. It's important to make sure that the ARCH process finishes its task well before another redo log group is filled. If the ARCH process can't keep up with redo log generation, Oracle will have to suspend database activity when all redo log groups have filled. Although database activity will resume as soon as ARCH can catch up and release a redo log group, users and batch jobs will most definitely notice the lack of system responsiveness. Tape ARCHIVELOG storage can be problematic Thankfully, most databases don't suffer from this affliction unless redo log devices are substantially faster than the devices storing archived redo logs. In particular, archiving redo logs directly to tape can result in the ARCH process not being able to keep up with redo log generation. Always archive redo logs to disk storage. (Disk is cheap!) You can check the time the ARCH process is using for each filled redo log. Do this by looking at the time of a log switch in the relevant ALERT.LOG file, and then comparing it to the timestamp of the resulting archived redo log file. If the time required for the ARCH process is close to the time between redo log switches, you should implement one or more of the following options: q Increase the size of redo logs. The ARCH process is more efficient when working with fewer and larger redo logs.
http://www.informit.com/content/0789716534/element_020.shtml (13 of 14) [26.05.2000 17:11:34]

informit.com -- Your Brain is Hungry. InformIT - Tuning Your Memory Structures and File Access From: Using Oracle8
q

Add more redo log groups. This will give the ARCH process more time to catch up during peak transaction rates without causing the database to suspend activity. Store archived logs on faster physical devices. By using RAID technology with high-speed disk drives, the ARCH process should be able to keep up with most any realistic redo log generation rate.

Designing Redo to Avoid LGWR and ARCH Contention

When a log switch occurs in Oracle, the LGWR process will immediately begin writing to the next redo log group in a circular fashion. Meanwhile, the ARCH process will begin reading the last redo log to make a copy in the archive log destination. Because these operations happen concurrently, each should be working with two distinct physical disk drives. Recall that when redo log group n is filled up, Oracle will always switch to the n+1 group (unless n is the last redo log group, in which case Oracle will start with group 1 again). Therefore, when redo log group n is full, a log switch occurs-log group n is read by the ARCH process, while LGWR moves on to writing in group n+1. Remember to isolate redo logs Because of the high I/O rates on redo log data files, remember to dedicate disk drives exclusively for redo log use. In an ideal system, at least four disk drives (two data files per group on different drives and two groups on completely different devices) will be dedicated for redo log use. To ensure that ARCH and LGWR are always working from different physical drives, odd-numbered redo log data files should always reside on completely different disk drives (and, ideally, different controllers) than even-numbered redo log data files. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_020.shtml (14 of 14) [26.05.2000 17:11:34]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Identifying and Reducing Contention

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

Identifying Rollback Segment Contention

r

Reducing Contention for Rollback Segments Identifying Contention for Dispatcher Processes Identifying Contention for Shared Server Processes

Identifying and Reducing Contention for Multithreaded Server Processes (MTS)

r r

q q

Identifying and Reducing Contention for Parallel Server Processes Identifying and Reducing Latch Contention
r r r r

Identifying and Reducing Contention for the LRU Latch Identifying and Reducing Contention for Space in Redo Log Buffers Identifying and Reducing Contention for Redo Log Buffer Latches Identifying and Reducing Contention for Library Cache Latches Reducing Contention for Free Lists

Identifying Free-List Contention

r

Identifying Rollback Segment Contention

Rollback segments are concurrently used by one or more transactions; any delays caused by contention on rollback segments affect performance. More transactions per rollback segment cause more contention and use space more efficiently, whereas fewer transactions per rollback segment cause less contention and waste more space. Interleaf rollback segments to reduce contention Interleaf the order of the rollback segments in the initialization file's ROLLBACK_SEGMENTS parameter so that the first is in one table-space, the next one is in the other, and so on. This will reduce the contention for rollback segments. Rollback segments hold transaction tables in their headers. Concurrent transactions contending for the same rollback segment will show up due to contention for the undo header. Each time a transaction begins, it's assigned to a rollback segment. This assignment can be done automatically or manually: q Automatic. When the first DDL or DML statement is issued, Oracle automatically assigns the transaction to a rollback segment. Queries are never assigned rollback segments. q Manual. You can use the SET TRANSACTION command with the USE ROLLBACK SEGMENT parameter and manually specify a rollback segment to use for a transaction. The rollback segment is

http://www.informit.com/content/0789716534/element_021.shtml (1 of 13) [26.05.2000 17:12:36]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

assigned for the duration of the transaction. By using this method, the developers can use the correct size rollback segment for a particular task. You should create rollback tablespaces to hold rollback segments. In a heavy transaction-based system, you should have at least two rollback tablespaces on separate disks. Rollback segment contention occurs when transactions request rollback segment buffers and those buffers are still busy with previous transaction rollback information. Rollback segment contention is reflected by contention for buffers that contain rollback segment blocks. V$WAITSTAT contains statistics for different classes of block. The following table shows the different classes of blocks tracked through this view for rollback information: Block class System Undo Header System Undo Block Undo Header Undo Block Description Buffers containing header blocks of the SYSTEM rollback segment Buffers containing blocks of the SYSTEM rollback segment other than header blocks Buffers containing header blocks of the rollback segments other than the SYSTEM rollback segment Buffers containing blocks (other than header blocks) of the rollback segments (other than the SYSTEM rollback segment)

Use the following queries to determine the number of requests for data and the number of waits for each class of block over a period of time. The following query gives you the total number of data requests: SELECT SUM (value) "DATA REQUESTS" FROM V$SYSSTAT WHERE name IN ('db block gets', 'consistent gets'); This query's output might look like this: DATA REQUESTS -----------------------223759 The following query provides the number of waits on rollback segments: SELECT class, count FROM V$WAITSTAT WHERE class LIKE '%undo%' AND COUNT > 0; This query's output might look like this: CLASS COUNT --------------------------system undo header 3145 system undo block 231 undo header 4875 undo block 774 As seen from these results, the number of waits for system undo header is (3145 / 223759)*100 = 1.4%, and the number of waits for undo header is (4875 / 223759)* 100 = 2.1%. Contention is indicated by the number of waits for any class being greater than 1 percent of the total number of requests. Contention is indicated by frequent occurrence of error ORA-01555, or when transaction table wait events are much greater than 0.

Reducing Contention for Rollback Segments

Increasing the number of rollback segments can reduce contention for rollback segments. After creating the new

http://www.informit.com/content/0789716534/element_021.shtml (2 of 13) [26.05.2000 17:12:36]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

rollback segments, you must reference them in the init.ora parameter ROLLBACK_SEGMENTS and also make them online. Use the following table as a guideline for determining the number of rollback segments to allocate, depending on the number of concurrent transactions anticipated. Number of concurrent transactions Fewer than 16 concurrent transactions Concurrent transactions between 16 and 32 More than 32 concurrent transactions Number of rollback segments 4 8 Number of transactions divided by 4

The following query will tell you whether the OPTIMAL setting for the rollback segments is appropriate: SELECT substr(name, 1,20), extents, rssize, aveactive, ave_shrink, extends, shrinks FROM v$rollname rn, v$rollstat rs WHERE rn.usn = rs.usn; This query's output might look like this: substr(name, 1, 20) extents rssize aveactive aveshrink extends shrinks -----------------------------------------------------------SYSTEM 4 207639 0 0 0 0 RB1 2 207489 0 0 0 0 RB2 4 207698 0 0 0 0 OPTIMAL is set properly if the average size of rollback segments is close to the size set for OPTIMAL; otherwise, you need to change the value for OPTIMAL. Also, if the value in the shrinks column is very large, that means that the OPTIMAL setting is set improperly. You can follow several other recommendations with respect to rollback segments to reduce contention: q Set NEXT to INITIAL. q Set MINEXTENTS equal to or greater than 20. q Set OPTIMAL to equal INITIAL multiplied by MINEXTENTS. q Set the value for INITIAL appropriately after determining the amount of undo generated by transactions, as follows: SELECT MAX(USED_UBLK) FROM v$transaction; Set INITIAL equal to or greater than MAX(USED_UBLK).

Use rollback segments wisely Assign large rollback segments to long-running queries and to transactions that modify a large amount of data. Avoid dynamic expansion and reduction of roll-back space. Also, don't create more rollback segments than your instance's maximum number of concurrently active transactions.

Identifying and Reducing Contention for Multithreaded Server Processes (MTS)

The components of the MTS configuration consist of the processes on the system, communication software, and the shared global area (SGA). A system using MTS has the following processes (see Figure 21.1): q A network listener process that connects user processes to dispatchers and dedicated servers. (This functionality is provided by Net8.) q One or more dispatcher processes. Dispatchers belong to a specific instance and are protocol-specific handlers. (Windows NT currently supports only the TCP/IP protocol.) q One or more server processes. Figure 21.1 : The components of the MTS configuration consist of the processes on the system, communication software, and the shared global area (SGA).

http://www.informit.com/content/0789716534/element_021.shtml (3 of 13) [26.05.2000 17:12:36]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

Use Net8 to connect to shared servers To use shared servers, a user process must connect through Net8, even if it's on the same machine as the Oracle instance. When you configure Oracle8 to use Multithreaded Server architecture, you have to deal with the contention for dispatcher and shared server processes. You can configure the number of dispatchers and server processes; there's no direct relationship between the number of dispatchers and shared servers. SGA enhancements include additional memory structures, such as request and response queues for handling service requests and returning responses to those requests. Session information is migrated from the PGA into the SGA (this section of the SGA is known as the user global area, or UGA), so that the correct response goes to the appropriate client. PGA in an MTS environment The program global area (PGA) of a shared-server process doesn't contain user-related data because this information needs to be accessible to all shared servers. The PGA of shared servers contains only stack space and process-specific variables. Session-related information is moved to the SGA, which should have enough space to store all session-specific information. How MTS server processes are run 1. When the listener is started, it starts listening on the listed addresses. It opens and establishes a communication path through which users connect to Oracle. The only services it's aware of are those defined in listener.ora. 2. When an Oracle instance configured for MTS is started, each dispatcher gets its random listen address and gives the listener this address, at which the dispatcher listens for connection requests. The dispatcher calls the listener by using the address specified in the init.ora parameter MTS_LISTENER_ADDRESS. 3. The listener adds the dispatcher's MTS_SERVICE and address to its list of known services. 4. The network listener process waits for incoming connection requests and determines whether a shared server process can be used. 5. If a dedicated server process is requested, it creates a dedicated server process and connects the user process to it; otherwise, it gives the user process the address of a dispatcher process with the lightest load. (Windows NT now supports only the TCP/IP protocol for MTS connections.) 6. The user process connects to the dispatcher and remains connected to it throughout the life of the user process. After the connection is established, the dispatcher creates a virtual circuit, which it uses to communicate with the shared servers. 7. The user process issues a request, which is placed by the dispatcher in the request queue in the SGA, where it's picked up by the next available shared server process. The request queue is common to all dispatchers. 8. The shared server process does all the necessary processing and returns the results to the response queue of the dispatcher in the SGA. Each dispatcher has its own response queue. 9. The dispatcher process returns the completed request to the user process. Because dispatchers have few responsibilities, each dispatcher can serve many clients, allowing a significantly high number of clients to be connected to the server.

Identifying Contention for Dispatcher Processes

Contention for dispatcher processes can be identified by the following symptoms: q High busy rates for dispatchers. The V$DISPATCHER view contains important statistics regarding the activity of dispatcher processes. Pay special attention to this view's [BUSY] and [IDLE] fields. q A steady increase in waiting time for responses in the dispatchers' response queues. The V$QUEUE view contains important statistics regarding the response queue activity of dispatcher processes. Pay special attention to this view's [WAIT] and [TOTALQ] fields Privileges needed for views By default, the V$DISPATCHER and V$QUEUE views are available only to the user SYS and to other users with the SELECT ANY TABLE system privilege. Examining Busy Rates for Dispatcher Processes

http://www.informit.com/content/0789716534/element_021.shtml (4 of 13) [26.05.2000 17:12:36]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

You can examine the efficiency of the dispatchers by using the following query over a period of time with the application running: View Code Four dispatchers are running on this instance. All have the status WAIT except one (D001), whose status is CONNECTED. This dispatcher is servicing a connect request from a client that will remain connected to this dispatcher for the lifetime of its session. None of these dispatchers have a high percentage of busyness; you could conclude that we could do with fewer dispatchers. You can use another query to determine the dispatcher that you need in the system: SELECT network "PROTOCOL", ( SUM(busy) / (SUM(busy) + SUM(idle)) ) * 100 "% TIME BUSY" FROM v$dispatcher GROUP BY network; PROTOCOL ========== decnet tcp % TIME BUSY ============ 0.5676849 61.4379021

From this result, you can see the following: q The DECnet dispatcher processes are busy nearly 0.5 percent of the time. q TCP dispatchers processes are busy nearly 61 percent of the time. Thus, you can conclude that there's contention for the TCP dispatchers. You can improve the performance by adding dispatchers for the TCP protocol. Examining Contention for Dispatcher Process Response Queues You can determine whether you have the correct number of dispatchers by finding the average wait time for the dispatchers of that protocol. You can use the following query for this purpose: SELECT network "PROTOCOL", DECODE( SUM(totalq), 0, 'No Responses', SUM(wait)/SUM(totalq) ) "Average wait time per response (1/100th of seconds)" FROM v$queue q, v$dispatcher d WHERE q.type = 'DISPATCHER' AND q.paddr = d.paddr GROUP BY network; This query will return the average time, in hundredths of a second, that a response waits in the response queue. A steady increase in the wait time would indicate that you need to add more dispatchers. Protocol -------decnet tcp Average wait time per response (1/100th of a second) --------------------------------------------------0.134180 235.38760

This result shows that the responses in the response queue of the DECnet dispatcher processes wait an average of 0.13 of a second, whereas the wait is much higher for the TCP dispatcher processes. Reducing Contention for Dispatcher Processes You have two options for reducing the contention of dispatchers: q Add dispatcher processes. You can dynamically adjust the number of dispatchers to improve performance. Use the ALTER SYSTEM command to change the number of dispatcher processes. q Enable connection pooling, a feature implemented with Net8 clients and dispatchers for MTS. It allows a limited number of physical connections to be shared among a large number of logical sessions. It uses a time-out mechanism to temporarily release an idle transport connection while maintaining its network session. Use of this feature is ideal when many clients run interactive applications such as email, which

http://www.informit.com/content/0789716534/element_021.shtml (5 of 13) [26.05.2000 17:12:36]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

have high idle time. Oracle8 provides an optional attribute POOL (or POO), which can be used with the parameter MTS_DISPATCHERS to enable the Net8 connection pooling feature in the init.ora file. The following example allows you to start the database with four TCP dispatchers and enables the Net8 connection pooling feature: MTS_DISPATCHERS = "(PROTOCOL=TCP) (DISPATCHERS=4) (POOL)" Limit the number of dispatchers The MTS_MAX_DISPATCHERS parameter determines the total number of dispatcher processes across all protocols. The default value of this parameter is 5; the maximum value is operating-system dependent.

Identifying Contention for Shared Server Processes

Contention for shared server processes can be determined by monitoring the wait times for requests in the request queue (V$QUEUE and V$SHARED_SERVERS). A steady increase in the wait time indicates that there's contention and that you need to increase the number of shared server processes. The query for examining the shared servers is similar to that of monitoring dispatchers but from the dynamic view V$SHARED_SERVERS: View Code Servers S000 and S002 are idle, whereas server S001 is very busy; its client could benefit from a dedicated server. The value of the REQUESTS column is cumulative since instance startup. You can use the following query to determine the average wait time for the requests in the request queue: SELECT DECODE( totalq, 0, 'No Requests', wait/totalq ) "Average wait time per requests (hundredths of seconds) " FROM v$queue WHERE type = 'COMMON'; Average wait time per requests (hundredths of seconds) ----------------------------------------------------------0.293734950 The result shows that there's an average wait time of 0.29 of a second in the request queue. Reducing Contention for Shared Servers Oracle8 automatically adjusts the number of shared servers to improve performance. You should try to identify requests that need to be serviced by dedicated servers and not shared servers. Limit the number of shared servers The MTS_MAX_SERVERS parameter determines the maximum number of shared server processes in the system. The default value of this parameter is 20; the maximum value is operating-system dependent. Theoretically, any application can be used with MTS, but in reality some applications will benefit from its usage, whereas others won't. Client/server applications that don't require a lot of processing from the client will benefit from the usage of MTS. An example of such an application is Oracle's own BUG database. During the day, bugs may be filed, updated, and closed by various people working in any number of departments. These transactions are short and don't tie up the server for an extended period of time, allowing the server to work on many requests simultaneously. However, not all the BUG database's functionality will benefit from MTS, such as running weekend reports and complex SELECT statements, which can tie up the shared servers. Suppose that an MTS instance has three dispatchers and two shared servers. If one client decides to submit such reports during the day, you are left with just one shared server, because the other will be tied up processing the reports. Now if a second client submits another such report, it will appear to other clients that the instance is locked up, because both shared servers are performing large reports. New requests won't be processed until a server becomes free. It will be better if the clients request a dedicated connection. As you can see, a mixed environment can usually give great benefits.

http://www.informit.com/content/0789716534/element_021.shtml (6 of 13) [26.05.2000 17:12:37]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

The following shows the init.ora file with MTS parameters: mts_service= "YOUR_SID" mts_listener_address="(ADDRESS=(PROTOCOL=tcp)(port=1521) (host=your_machine))" mts_dispatchers= "tcp, 1" mts_max_dispatchers=10 mts_max_servers=10 mts_servers=4

Identifying and Reducing Contention for Parallel Server Processes

You can use the dictionary view V$PQ_SYSSTAT for determining the appropriate number of parallel server processes for an instance: SELECT STATISTIC, VALUE FROM v$pq_sysstat WHERE statistic = "Servers Started " or statistic = "Servers Shutdown"; STATISTIC VALUE -----------------------Servers Started 48 Servers Shutdown 45 The following statistics can be helpful in identifying the contention for parallel server processes: q Servers Busy q Servers Idle q Servers Started q Servers Shutdown Use the following query to determine whether the set numbers of parallel server processes are really busy: SELECT STATISTIC, VALUE FROM v$pq_sysstat WHERE statistic = "Servers Busy" STATISTIC VALUE -----------------------Servers Busy 65 If the query's result shows that the number of busy parallel server processes is typically less than PARALLEL_MIN_SERVERS, you should consider decreasing the value of PARALLEL_MIN_SERVERS in initsid.ora. If servers are continuously starting and shutting down, however, consider increasing PARALLEL_MIN_SERVERS. This parameter should be set to the number of concurrent parallel operations multiplied by the average number of parallel server processes used by a parallel operation. The maximum parallel servers should not exceed your machine capacity The maximum number of parallel server processes for an instance depends heavily on the capacity of your CPUs and your I/O bandwidth. Set PARALLEL_MAX_SERVERS to the maximum number of concurrent parallel server processes that your machine can manage.

Identifying and Reducing Latch Contention

Oracle uses different types of locking mechanisms: q Latches q Enqueues q Distributed locks
http://www.informit.com/content/0789716534/element_021.shtml (7 of 13) [26.05.2000 17:12:37]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8
q

Global locks (used in parallel instance implementations)

You can't control what latches to use and when to use them, but you can adjust certain init.ora parameters to tune Oracle to most efficiently use latches and reduce latch contention. Latches protect internal data structures by maintaining a defined method of accessing them. If a process can't obtain a latch immediately, it waits for the latch, resulting in a slowdown and additional CPU usage. The process, in the meantime, is "spinning." Latch contention occurs when two or more Oracle processes attempt to obtain the same latch concurrently. You can detect latch contention by using the V$LATCH, V$LATCHHOLDER, and V$LATCHNAME data dictionary views. The following queries can be used to provide latch information: q Obtain the name of a latch from the latch address: SELECT name FROM v$latchname ln, v$latch l WHERE l.addr = '&addr' AND l.latch# = ln.latch# ; Obtain systemwide latch statistics: SELECT ln.name, l.addr, l.gets, l.misses, l.sleeps, l.immediate_gets, l.immediate_misses, lh.pid FROM v$latch l , v$latchholder lh , v$latchname ln WHERE l.addr = lh.laddr (+) AND l.latch# = ln.latch# ORDER BY l.latch# ; Display statistics for any latch X: SELECT ln.name, l.addr, l.gets, l.misses, l.sleeps, l.immediate_gets, l.immediate_misses, lh.pid FROM v$latch l , v$latchholder lh , v$latchname ln WHERE l.addr = lh.laddr (+) AND l.latch# = ln.latch# AND ln.name like '%X%' ORDER BY l.latch# ; The following table lists all the latches that are of concern to Oracle DBAs: Latch Number 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Name Latch wait list Process allocation Session allocation Session switching Session idle bit Messages Enqueues Trace latch Cache buffers chain Cache buffers LRU chain Cache buffer handles Multiblock read objects Cache protection latch System commit number Archive control Redo allocation Redo copy Instance latch Lock element parent latch

http://www.informit.com/content/0789716534/element_021.shtml (8 of 13) [26.05.2000 17:12:37]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

DML lock allocation Transaction allocation Undo global data Sequence cache Sequence cache entry Row cache objects Cost function User lock Global transaction mapping table Global transaction Shared pool Library cache Library cache pin Library cache load lock Virtual circuit buffers Virtual circuit queues Virtual circuits Query server process Query server free lists Error message lists Process queue Process queue reference Parallel query stats

The cache buffers chains latch is needed when user processes try to scan the SGA for database cache buffers. Adjusting the DB_BLOCK_BUFFERS parameter can reduce contention for this latch. The cache buffers LRU chain latch is needed when user processes try to scan the LRU chain containing all the dirty blocks in the buffer cache. Increasing the DB_BLOCK_BUFFERS and DB_BLOCK_WRITE_BATCH parameters can reduce contention for this latch. The row cache objects latch is needed when user processes try to access the cached data dictionary values. Tuning the data dictionary cache can reduce contention for this latch. Increasing the size of the shared pool (SHARED_POOL_SIZE) can be used to achieve this result.

Identifying and Reducing Contention for the LRU Latch

The least recently used (LRU) latch controls the replacement of buffers in the buffer cache. Contention for the LRU latch can be identified by querying V$LATCH, V$SESSION_EVENT, and V$SYSTEM_EVENT. V$LATCH shows cumulative values since instance startup. Design your applications to reduce LRU contention Application design can have a significant influence on LRU contention because the design will ultimately determine the queries and DML statements that in turn can affect contention for LRU latches. The initialization parameter DB_BLOCK_LRU_LATCHES specifies the maximum number of LRU latches on your system. Each LRU latch controls a set of buffers. You can use the following guidelines to determine the value for DB_BLOCK_LRU_LATCHES: q DB_BLOCK_LRU_LATCHES can range from 1 to twice the number of CPUs in the system. q A latch should have at least 50 buffers in its set. q Oracle uses only one LRU latch in a single-process mode. LRU latches for SMP and non-SMP machines For SMP machines, Oracle automatically sets the number of LRU latches to be one half the number of CPUs on the system. One LRU latch is sufficient for non-SMP machines.

http://www.informit.com/content/0789716534/element_021.shtml (9 of 13) [26.05.2000 17:12:37]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

Identifying and Reducing Contention for Space in Redo Log Buffers

Redo information must be written to the redo log before a transaction is completed. Any kind of bottleneck in the redo log can cause the performance of all the processes in the system to be affected. You should therefore check for contention of the redo log buffers and the redo log buffer latches. The dynamic performance table V$SYSSTAT stores systemwide statistics that can be used to identify contention for various resources. The statistic REDO BUFFER ALLOCATION RETRIES reflects the number of times a user process waits for space in the redo log buffer. The process may have to wait if the LGWR process isn't fast enough to write the redo entries from the redo log buffer to a redo log file. These statistics are available only to the SYS user or users with the SELECT ANY TABLE privilege. You can use the following query to determine contention for space in the redo log buffer: SELECT name, value FROM V$SYSSTAT WHERE name = 'redo buffer allocation retries'; The value from this query must be almost 0. If it increases consistently, contention is indicated. You have several options for reducing the contention for space in the redo log buffer: q Increase the size of the redo log buffer by increasing the LOG_BUFFER parameter. LOG_BUFFER (in bytes) must be a multiple of DB_BLOCK_SIZE. q Improve the checkpointing process. q Improve the archiving process.

Identifying and Reducing Contention for Redo Log Buffer Latches

The V$LATCH dynamic performance view contains statistics for the activity of the various latches. A process can request a latch in one of two ways: q Willing to wait. If the requested latch is unavailable, the requesting process will wait for some time and request the latch again. This wait-request cycle continues until the latch becomes available. The following V$LATCH columns reflect willing-to-wait requests: GETS MISSES SLEEPS
q

Number of successful willing-to-wait requests for a latch Number of unsuccessful, initial willing-to-wait requests for a latch Number of times a process waited and requested a latch after an initial request

Immediate. If the requested latch is unavailable, the requesting process doesn't wait but continues processing. The following V$LATCH columns reflect immediate requests: IMMEDIATE GETS IMMEDIATE MISSES Number of successful immediate requests for a latch Number of unsuccessful immediate requests for a latch

The following query ca an be used to monitor contention of the redo allocation and the redo copy latch: SELECT ln.name, gets, misses, immediate_gets, immediate_misses FROM V$LATCH l, V$LATCHNAME ln WHERE ln.name IN (' redo allocation', 'redo copy') AND ln.latch# = l.latch#; This query's output might look like this: NAME GETS MISSES IMMEDIATE_GETS IMMEDIATE_MISSES -------------- ----- ------- --------------- --------------redo alloc... 12580 215 5 0 redo copy 12 0 1223 2 Contention exists for a latch if either of the following is true: q Ratio of MISSES to GETS exceeds 1 percent q Ratio of IMMEDIATE_MISSES to the sum of IMMEDIATE_MISSES and IMMEDIATE_GETS exceeds

http://www.informit.com/content/0789716534/element_021.shtml (10 of 13) [26.05.2000 17:12:37]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

1 percent The example shows a redo allocation latch contention; the ratio of misses to gets is 1.7 percent. The redo allocation latch controls space allocation for redo entries in the redo log buffer. An Oracle process must obtain the redo allocation latch before allocating space in the redo log buffer. There's only one redo allocation latch; therefore, only one process can allocate space in the redo log buffer at a time. Latch contention doesn't occur on single-CPU machines Only one process can be active at a given time on a single-CPU machine; therefore, latch contention rarely occurs. You can reduce contention for this latch by minimizing copying on it, which in turn reduces the time that any single process holds the latch. To do so, decrease the value of the LOG_SMALL_ENTRY_MAX_SIZE parameter, which determines the number and size of redo entries copied on the redo allocation latch. Whereas the redo allocation latch is held only for a short period of time, the redo copy latch is held for a longer amount of time because the user process first obtains the redo copy latch, and then the redo allocation latch. The process performs allocation and then releases the allocation latch. The copy is then performed under the copy latch, after which the redo copy latch is released. On multiple CPU machines, the LOG_SIMULTANEOUS_COPIES parameter determines the number of redo copy latches. Multiple redo copy latches allow multiple processes to concurrently copy entries to the redo log buffer. The default value of this parameter is the number of CPUs available to the instance; the maximum value is twice the number of CPUs. To reduce contention, increase the value of LOG_SIMULTANEOUS_COPIES. Another way in which you can reduce redo copy latch contention is by prebuilding the redo entry before requesting the latch. The LOG_ENTRY_PREBUILD_THRESHOLD parameter can be set to achieve this result. The default value for this parameter is 0. When this parameter is set, any redo entry of a smaller size than this parameter must be prebuilt.

Identifying and Reducing Contention for Library Cache Latches

The library cache latches are required to prevent multiple access to a shared library cache entry. There are three types of library cache latches: q The library cache latch-the highest latch level-is needed before getting a lock on a handle. q The pin latch is needed by a process before pinning a heap. q The load-lock latch is needed to load a library cache entry. The following query can be used to identify library cache latch contention: SELECT count(*) number_of_waiters FROM v$session_wait sw, v$latch l WHERE sw.wait_time = 0 AND sw.event = 'latch free' AND sw.p2 = l.latch# AND l.name like 'library%'; In this query, a large number of waiters indicates that you have contention. You can reduce contention for library cache latches in the following ways: q Minimize shared pool fragmentation. q Increase the sharing of statements, which in turn decreases the completed reloads. q Reduce the amount of parsing that's going on in the system. You can use the following query to identify the SQL statements that are receiving a lot of parse calls: SELECT sql_text, parse_calls, executions FROM v$sqlarea WHERE parse_calls > 100 AND executions < 2*parse_calls; Set CURSOR_SPACE_FOR_TIME to TRUE. Use fully qualified table names.

q q

http://www.informit.com/content/0789716534/element_021.shtml (11 of 13) [26.05.2000 17:12:37]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

Identifying Free-List Contention

Oracle maintains information in memory that allows instantaneous access to information on disk. Some of this information keeps track of the space available in the tablespaces for inserting new rows. You can indicate at table-creation time how much information to keep in memory for tracking available blocks for creating new records. This is the free list for the table. The following code creates a table with 12 free lists: create table test_table (id number, desc varchar2(20)) freelist 12; Free-list contention is generally indicated by excessive disk I/O for block ID requests to create new rows. This contention occurs when multiple processes are searching the free list at the same time. The dynamic view V$WAITSTAT contains statistics for block contention. Free-list contention is reflected by contention for free data blocks in the buffer cache. You can use the following query to determine what percentage of the total requests was actually waiting for free blocks: SELECT class,count FROM V$WAITSTAT WHERE class = 'free list'; This query's output might look like this: CLASS COUNT ----------------free list 39 You can use the following query to determine the number of requests for DB_BLOCK_BUFFERS and consistent reads: SELECT name, value FROM v$sysstat WHERE name in ('db block gets', 'consistent gets'); This query's output might look like this: NAME VALUE ---------------------db block gets 14236 consistent gets 10437 By using the preceding query, you can determine the number of waits for free lists: free list wait events = ( (free list count) / (db block gets + consistent gets)) * 100 = (39/ (14236 + 10437 ) * 100 = 0.15 Identify free-list contention 1. Query V$WAITSTAT for contention on DATA BLOCKS. 2. Query V$SYSTEM_EVENT and determine BUFFER BUSY WAITS. 3. For each buffer-busy wait, use V$SESSION_WAIT to determine the FILE, BLOCK, and ID. 4. The following query shows the segment_name and segment_type for the objects and free lists with buffer busy waits: SELECT segment_name "segment", segment_type "type" FROM dba_extents WHERE FILE_ID = file AND BLOCK BETWEEN block_id and block_id + blocks;
http://www.informit.com/content/0789716534/element_021.shtml (12 of 13) [26.05.2000 17:12:37]

informit.com -- Your Brain is Hungry. InformIT - Identifying and Reducing Contention From: Using Oracle8

5. You can use the segment and type from the preceding query in the following query to determine the free lists having contention: SELECT FREELISTS FROM DBA_SEGMENTS WHERE SEGMENT_NAME = segment AND SEGMENT_TYPE = type; If the number of free-list wait events is greater than 1 percent of the total number of requests, you have contention for free list.

Reducing Contention for Free Lists

Recreating the table with a larger value of the FREELISTS storage parameter can reduce contention for free lists of a table. There are alternative ways for recreating the table that has free-list contention. The first example uses the table TEST. Using Export/Import to reduce free-list contention 1. Export the table with the grants and indexes. 2. Drop the table. 3. Recreate the table with an increased value for FREELISTS. 4. Import the table with IGNORE=y. You should have at least enough free lists to accommodate the number of concurrent processes that will be inserting into the table. Using CTAS (Create Table as SELECT) to reduce free-list contention 1. Create a new table with the desired value for FREELISTS and the same structure as the old table. 2. Select data from the old table into the new table. 3. Drop the old table. 4. rename the new table. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_021.shtml (13 of 13) [26.05.2000 17:12:37]

informit.com -- Your Brain is Hungry. InformIT - Tuning for Different Types of Applications From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Tuning for Different Types of Applications

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

Transaction Processing Tuning

r r r

Tuning Rollback Segments Using Discrete Transactions Transaction Processing Monitors (TPMs) Adding Indexes Managing Sort Space Managing Hash Join Space Designing Tables for Star Queries Parallel Operations

DSS and Data-Warehousing Tuning

r r r r r

Transaction Processing Tuning

Online Transaction Processing (OLTP) systems have several characteristics: q High throughput q Insert/update intensive q Large amount of concurrent users q Data volume changing rapidly Typical OLTP applications include airline reservation systems and banking applications. When you design an OLTP system, you must ensure that the large numbers of concurrent users don't affect the system performance. At the same time, you should get high availability, speed, and recoverability. You need to consider several issues when using OLTP systems: q Rollback segments q Indexes, clusters, and hashing q Transaction modes q Size of the data block q Transaction processing monitors q Multithreaded server q Well-tuned memory structures

Tuning Rollback Segments

In an OLTP environment, an increased number of transactions occur, which may require a lot of rollback

http://www.informit.com/content/0789716534/element_022.shtml (1 of 10) [26.05.2000 17:13:25]

informit.com -- Your Brain is Hungry. InformIT - Tuning for Different Types of Applications From: Using Oracle8

segments. When you're using parallel DML statements, the rollback segments become more important. You should have rollback segments that belong to tablespaces with lots of free space. You should also have unlimited rollback segments or a very high value for MAXEXTENTS. SEE ALSO More information on tuning rollback segments,

Using Discrete Transactions

Discrete transactions can be used concurrently with standard transactions. Discrete transactions are useful for transactions with certain characteristics: q Transactions are short and non-distributed. q Transactions modify very few database blocks. q Transactions don't change individual database blocks more than once per transaction. q Transactions don't modify data that will be needed by long-running queries for read consistency. q Transactions don't need to see the new value of data after modifying the data. q Modified tables don't contain any LONG values. ORA-1555 during discrete transactions Undo information isn't generated during a discrete transaction. Therefore, you may get "snap-shot too old" errors if a discrete transaction starts and commits during a long query, and the discrete transaction modifies blocks used by the long query. When you use a discrete transaction, all the changes made to any data will be deferred until the transaction commits. Redo information is generated but undo information isn't generated. The redo information is stored in a separate location in memory and gets written to the redo log when the transaction commits. Changes made during discrete transactions Discrete transactions can't see their own changes. Therefore, they can't perform inserts or updates on both tables involved in a referential integrity constraint. Use discrete transactions 1. Set the parameter DISCRETE_TRANSACTIONS_ENABLED to TRUE in the initialization file. (If DISCRETE_TRANSACTIONS_ENABLED is set to FALSE, all the transactions will be run as standard transactions.) 2. Use the procedure BEGIN_DISCRETE_TRANSACTION as the first statement in a transaction. Scope of a discrete transaction The procedure BEGIN_DISCRETE_TRANSACTION is effective only for the current transaction. As soon as the transaction commits or rolls back, the next transaction will be treated as a standard transaction. The example in Listing 22.1 uses a discrete transaction for a movie theater application. The name of the movie, show time, and number of tickets are passed as the arguments to this procedure. The procedure checks the tickets available. If enough tickets are available, it issues the tickets; otherwise, it will indicate the number of tickets available for that movie and show time. LISTING 22.1 A movie theater application using a discrete transaction View Code Discrete transaction failure In this example, if the discrete transaction fails, the DISCRETE_TRANSACTION_FAILED exception occurs and the transaction is rolled back. The second iteration of the loop will re-execute the transaction but this time as a standard transaction.

Transaction Processing Monitors (TPMs)

When a large number of concurrent users are connecting to the database, the demand for resources can become very heavy. On UNIX, the limiting factor is usually the physical memory. On Windows NT, each process can address a maximum of 4GB address space; 2GB of this 4GB address space is shared between processes. The

http://www.informit.com/content/0789716534/element_022.shtml (2 of 10) [26.05.2000 17:13:25]

informit.com -- Your Brain is Hungry. InformIT - Tuning for Different Types of Applications From: Using Oracle8

shadow processes are implemented as threads within a single process; therefore, the limiting factor becomes the 2GB address space limit. There are several solutions to this problem, such as Multi-Threaded Server, concurrent manager, and connection multiplexing. Fault tolerance using TPM The application client isn't aware of the application server servicing it; therefore, fault tolerance can be provided by using the transaction monitor architecture. Another approach is to split the application logic into an application client and application server. The application client becomes responsible for the collection and presentation of the data, whereas the application server becomes responsible for providing business services that it implements through accessing various data stores (resource managers). The application server will normally be written in a 3GL such as C or COBOL, and the interface to the resource manager is using the normal precompiler or API mechanisms. This environment is normally managed with a transaction processing monitor (also called a transaction monitor), such as BEA Tuxedo. The TPM will normally provide the following: q A messaging interface q Message routing between client and server q Application server management including registration, startup, shutdown, and load balancing A typical application server's logic 1. Connect to Oracle: EXEC SQL CONNECT 2. Wait for the message from the application client (an ATMI call for BEA Tuxedo). 3. Access Oracle: EXEC SQL SELECT EXEC SQL UPDATE 4. Commit: EXEC SQL COMMIT; 5. Send a reply to the application client (an ATMI call for BEA Tuxedo). 6. Return to step 2 and repeat the process. In a high-concurrency environment, many application clients can share the same application server. The TPM is responsible for routing messages. As the load increases, the TPM can also spawn more application servers. Monitoring transactions in pending state The DBA_2PC_PENDING and the DBA_2PC_NEIGHBORS views can be used to monitor if the transaction is in a pending state. Distributed transaction processing can be obtained in the transaction-processing (TP) architecture by routing messages to different application servers. This can be homogeneous (Oracle-Oracle) or heterogeneous (Oracle-Sybase). The limitation is that the transaction is committed separately on the data stores. Thus, you'll need to use XA to achieve global heterogeneous transactions. In the XA architecture, each resource manager exports an XA API. The TPM becomes the 2PC coordinator and uses the XA API to control the prepare/commit/rollback of transactions within the resource managers. Direct commits and rollbacks from the application servers are replaced with calls to the TPM through the XA interface. Transaction Recovery in the XA Environment If the TPM crashes or becomes unavailable while performing a 2PC, a transaction may be left in pending state. In this state, it may be holding locks and preventing other users from proceeding. You can use the commit force or rollback force statements as needed to manually force the transaction to commit or rollback. Purchasing a TPM Several vendors provide UNIX TPMs. The following support XA: q Tuxedo System/T, UNIX System Laboratories q Top End, NCR Corporation q Encina, Transarc Corporation

http://www.informit.com/content/0789716534/element_022.shtml (3 of 10) [26.05.2000 17:13:25]

informit.com -- Your Brain is Hungry. InformIT - Tuning for Different Types of Applications From: Using Oracle8
q q

CICS/6000, IBM Corporation CICS 9000, Hewlett-Packard

The following don't currently support XA: q VIS/TP, VISystems, Inc. q UniKix, UniKix, Inc. q Micro Focus Transaction System, Micro Focus Writing an Oracle TPM Application The TPM vendor documentation should describe the actual APIs used to talk to the TPM. However, all of them have a way to indicate where a transaction begins and ends and a way to send a request and receive a response from a client to a server. In the following example, which uses Tuxedo /T verbs, the SQL COMMIT is replaced with the TPM COMMIT. The first example is an Oracle managed transaction. For the client, use Tpcall("debit_credit"); Use the following for the server: Debit_credit_service (TPSCVCINFO *input) { extract data from the input; EXEC SQL UPDATE debit_data; EXEC SQL UPDATE credit_data; EXEC SQL COMMIT WORK; Tpreturn(output_data); } The next example is a TPM-managed transaction using XA. Use the following for the client: Tpbegin(); Tpcall("debit"); Tpcall("credit"); Tpcommit(); For server 1, use Debit_service(TPSCVCINFO *input) { extract data from the input; EXEC SQL UPDATE debit_data; Tpreturn(output_data); } For server 2, use the following: credit_service(TPSCVCINFO *input) { extract data from the input; EXEC SQL UPDATE credit_data; Tpreturn(output_data); } For the TPM API and programming overview, you can obtain the material from the vendor.

DSS and Data-Warehousing Tuning

DSS systems generally perform queries on huge amounts of data that has been gathered from OLTP systems. You need to consider several issues when using DSS systems: q Index management q Data block size

http://www.informit.com/content/0789716534/element_022.shtml (4 of 10) [26.05.2000 17:13:25]

informit.com -- Your Brain is Hungry. InformIT - Tuning for Different Types of Applications From: Using Oracle8
q q

Star queries Parallel execution

Adding Indexes
An index is generally used to provide a fast access path to the data. When using indexes in a DSS environment where you would be dealing with a huge amount of data, you need to take several special measures: q Create indexes after inserting data in the table. Use SQL*Loader or Import to load the data first, and then create the index. This will be faster because the index doesn't have to be maintained for each insertion; instead, it's created at the end when all the data has been inserted. q Create enough indexes based on the type of queries you'll be running. Using the UNRECOVERABLE option Using the UNRECOVERABLE option during index creation can speed up index creation. Because this won't generate any redo log records during index creation, you should back up after the index is created. In the DSS system, the data changes won't be a lot, so you should be able to create relatively more indexes without worrying about the performance. In general, be careful when creating indexes because unnecessary indexes can degrade performance. When you're using star queries, the indexes on the fact table can be partitioned or non-partitioned. Local partitioned indexes are the simplest, but their disadvantage is that a search of local non-prefixed index requires searching of all the index partitions.

Managing Sort Space

In a DSS environment, you usually will end up with queries that require some kind of sorting. Several initialization parameters can be manipulated to manage the amount of sort space used by these operations: q SORT_AREA_SIZE. You can set this parameter from the initialization file. It specifies the amount of memory to allocate per parallel server process for sort operations. If you have sufficient memory on your system, you'll benefit in performance by setting a large value for SORT_AREA_SIZE because the entire operation can be performed in memory. If memory isn't enough, you should set SORT_AREA_SIZE to a lower value and increase the size of the buffer cache so that blocks from temporary sort segments can be cached. The SORT_AREA_SIZE is relevant to parallel query operations and to the query portion of DML or DDL statements. Effect of sort area size As the sort area size decreases, the amount of I/O increases due to the need to perform merges on a large number of runs. A large sort area, on the other hand, can result in a high operating system paging rate.
q

SORT_DIRECT_WRITES. You can set this parameter in the initialization file. The recommended value is AUTO. When this parameter is set to AUTO and SORT_AREA_SIZE is greater than 10 times the buffer size, the buffer cache will be bypassed for the writing of sort runs. Avoiding the buffer cache can improve performance by reducing the path length, reducing memory bus utilization, and reducing LRU latch contention on SMP machines. SORT_DIRECT_WRITES has no effect on hashing. SORT_AREA_RETAINED_SIZE. This parameter specifies the maximum amount of User Global Area (UGA) memory retained after a sort run completes. If more memory is required by a sort, a temporary segment is allocated and the sort becomes an external sort. SORT_AREA_RETAINED_SIZE is maintained for each sort operation in a query.

Large sort areas can be used effectively by combining a large SORT_AREA_SIZE with a minimal SORT_AREA_RETAINED_SIZE. Release memory after completing sorts The SORT_AREA_RETAINED_SIZE parameter allows you to specify the level to which memory should be released as soon as possible after a sort completes. If memory isn't released until the user disconnects, large sorts will create problems in the system.

Managing Hash Join Space

If Cx(Y) is used to represent the cost to perform operation x on table Y, the cost to perform a hash join is as

http://www.informit.com/content/0789716534/element_022.shtml (5 of 10) [26.05.2000 17:13:25]

informit.com -- Your Brain is Hungry. InformIT - Tuning for Different Types of Applications From: Using Oracle8

follows (provided that you have sufficient memory available): Chj(T1,T2) < Cread(T1) + Cread(T2) + Chash(T1,T2) When you use hash join operations, an in-memory hash table is built from smaller table, and then the larger table is scanned and joined by using a hash table probe. Hash joins are applicable to equijoins, anti-joins, and outer joins. Indexes aren't required to perform a hash join. Suppose that you have two tables, french and engineers, and want the names of all the French engineers. french is the smaller of the two tables. SELECT /*+ use_hash(french) */ french.name FROM french, engineers WHERE french.name = engineers.name To perform the hash join, an area of memory known as the hash memory is allocated (see Figure 22.1). Figure 22.1 : The in-memory hash table is obtained by using the french table. Applying hash functions to data When applying hash functions to data, Key1 = Key2 => hash Key1 = hash(Key2 ) but the reverse isn't true. (hash represents a hash function, and Key1and Key2represent data.) The first stage of the join involves scanning and partitioning the french table and building an in-memory hash filter and hash table. This is the build, and the french table is known as the build input. The french table is hash partitioned into smaller chunks so that at least one partition can be accommodated in the hash memory, which reduces the number of comparisons during the join. As the build input is scanned, parts of some partitions may be written to disk (temporary segment). A hash filter is created for each partition and stays in memory even if the partition doesn't fit. The hash filter is used to efficiently discard rows that don't join. After the french table is scanned completely, the size of each partition is known, as many partitions as possible are loaded into memory, and a single hash table is built on the in-memory partitions. Case 1: The hash memory is large enough for all the partitions of the french table; therefore, the entire join is completed by simply scanning the engineers table and probing the build. Case 2: The hash memory isn't large enough to fit all the partitions of the french table. In this case, the engineers table is scanned, and each row is partitioned using the same method (see Figure 22.2). Then for each row, q The "no hope" rows are discarded after verifying them with the hash filter. q Rows that correspond to in-memory french partitions are joined immediately, and rows that don't are written to an engineers partition on disk (temporary segment). Figure 22.2 : Hash joins are performed by using the in-memory hash table and the hash filters. After the engineers table is scanned, phase 2 begins. In phase 2, the smaller of the french and engineers partitions are scanned into memory and a hash table is built. The larger partition is scanned, and the join is completed by probing the hash table. If a partition won't fit in memory, the join will degenerate to a nested loop type mechanism. You can set the following parameters in the initialization file or by using the ALTER SESSION command from Server Manager: q HASH_JOIN_ENABLED (the default is TRUE). Setting this to TRUE allows you to use hash joins. q HASH_AREA_SIZE (the default is twice the SORT_AREA_SIZE). This parameter specifies the size of the hash memory. It should be approximately half of the square root of S, where S is the size (in MB) of the smaller of the inputs to the join operation. The value shouldn't be less than 1MB. HASH_AREA_SIZE is relevant to parallel query operations and to the query portion of DML or DDL statements. Set the size of hash area appropriately Each process that performs a parallel join operation uses an amount of memory equal to HASH_AREA_SIZE. Setting HASH_AREA_SIZE too large can cause the system to run out memory, whereas a setting that's too small can degrade performance.

http://www.informit.com/content/0789716534/element_022.shtml (6 of 10) [26.05.2000 17:13:25]

informit.com -- Your Brain is Hungry. InformIT - Tuning for Different Types of Applications From: Using Oracle8
q

HASH_MULTIBLOCK_IO_COUNT (the default is DB_FILE_MULTIBLOCK_IO_COUNT). This parameter specifies the number of sequential blocks a hash join should read and write in a single I/O.

Designing Tables for Star Queries

Star queries are most suitable for data warehouses, where there's a huge amount of data but relatively few data changes. A star schema is a natural data representation for many data warehousing applications; it consists of one very large table, referred to as a fact table, and many smaller tables, called dimension tables. (Note that the rule-based optimizer doesn't recognize star schemas.) Dimension tables aren't related to each other; however, each dimension table is related to the fact table through a primary key/foreign key relationship. Usually, there's a b*tree concatenated index on the columns of the fact table that are related to the dimension tables. An example of a star schema would be a retail environment (see Figure 22.3). Figure 22.3 : In a star schema, the relationship between the involved tables represents a structure. The following tables can be part of a retail environment: q sales with columns customer_key, supplier_key, part_key, store_key, date, quantity, and total_price q suppliers with columns supplier_key, name, address, and telephone q customers with columns customer_key, name, and address q parts with columns part_key, name, and cost q stores with columns store_key, address, and telephone The sales table can be the fact table and contain several millions of records for the different sales transactions. The dimension tables-suppliers, customers, parts, and stores-are relatively small and provide additional details about a sale. If you need to find sales information, use a star query. It uses specific details from the dimension tables. For example, if you want to find the total sales of a specific part purchased by a specific group of customers from a specific store and sort the results by suppliers, use the following query: Select supplier.name, sum(total_sales) from sales, customers, parts, suppliers, stores where sales.customer_key = customer.customer_key and sales.part_key = parts.part_key and sales.supplier_key = suppliers.supplier_key and sales.store_key = stores.store_key and customers.name in ('IBM','HP','COMPAQ') and parts.name = '2 GB disk' and stores.name = 'EGG HEAD SOFTWARE' group by suppliers.name; Tuning Star Queries To efficiently use star queries, you must use the cost-based optimizer (set OPTIMIZER_MODE to CHOOSE) and analyze with COMPUTE STATISTICS all the tables involved in the star query. To analyze tables, use the following command from the Server Manager prompt: SVRMGR> ANALYZE TABLE tablename COMPUTE STATISTICS; You can also tune star queries in other ways: q You can improve the optimizer plan by using hints. The most effective method is to order the tables in the FROM clause in the order of the keys in the index with the large table last. Then use the USE_NL hint or the STAR hint. STAR hint versus USE_NL hint The STAR hint /*+ STAR */ forces a star query plan to be used if possible. The USE_NL hint /* + USE_NL (table) */ causes each specified table to join with another row source with a nested loops join, using the specified table as the inner table.
q

The columns in the concatenated index of the fact table should take advantage of any ordering of the data. If all the queries specify predicates on each dimension table, a single concatenated index is sufficient, but

http://www.informit.com/content/0789716534/element_022.shtml (7 of 10) [26.05.2000 17:13:25]

informit.com -- Your Brain is Hungry. InformIT - Tuning for Different Types of Applications From: Using Oracle8

if you use queries that omit the leading columns of the concatenated index, you'll have to create additional indexes. Denormalized views can be effective at times when too much normalization of information can cause the optimizer to consider many permutations and result in very slow queries. For example, you have two tables, brands and manufacturers, that can be combined into a product view as follows: CREATE VIEW product AS SELECT /*+ NO_MERGE */ * FROM brands,manufacturers WHERE brands.mfkey = manufacturers.mfkey;

This will improve performance by caching the result of the view and reducing the executions of the small table joins. You also can use a star transformation by setting STAR_TRANSFORMATION_ENABLED to TRUE in the initialization file and using the STAR_TRANSFORMATION hint in the query. Tables with the following characteristics can't be used with a star transformation, however: q Tables with very few bitmap indexes q Anti-joined tables q Tables already used as dimension tables in a subquery q Remote tables q Tables that are unmerged views The star transformation is ideal under any of the following conditions: q The fact table is sparse. q There are a lot of dimension tables. q In some queries, not all dimension tables have constraining predicates. The star transformation doesn't rely on computing a Cartesian product of the dimension tables; instead, it uses bitmap indexes on individual fact table columns. It works by generating new subqueries that can be used to drive a bitmap index access path for the fact table. SEE ALSO More information on using the cost-based optimizer, See how to use hints, More information on bitmap indexes,

Parallel Operations
Oracle can perform the following operations in parallel: q Parallel query q Parallel DML (INSERT, UPDATE, DELETE, APPEND hint, and parallel index scans) q Parallel DDL q Parallel recovery q Parallel loading q Parallel propagation (for replication) You should try to parallelize operations that have high elapsed time or process a large number of rows. Tuning Parallel Operations in a DSS Environment You can use several techniques to optimize parallel operations. One way is to increase the default degree of parallelism for I/O- bound operations and decrease the degree of parallelism for memory-bound operations. Follow these guidelines when adjusting the degree of parallelism: q Use the ALTER TABLE command or hints to change the degree of parallelism. q Reducing the degree of parallelism will increase the number of concurrent parallel operations. q If the operation is I/O-bound, spread the data over more disks than there are CPUs and then increase the parallelism in stages until it becomes CPU-bound. You also can verify that all the parts of the query plan for SQL statements that process huge amounts of data are executing in parallel. By using EXPLAIN PLAN, verify that the plan steps have an OTHER_TAG of PARALLEL_TO_PARALLEL, PARALLEL_TO_SERIAL, PARALLEL_COMBINED_WITH_PARENT, or

http://www.informit.com/content/0789716534/element_022.shtml (8 of 10) [26.05.2000 17:13:25]

informit.com -- Your Brain is Hungry. InformIT - Tuning for Different Types of Applications From: Using Oracle8

PARALLEL_COMBINED_WITH_ CHILD. Any other keyword or null indicates serial execution and possible bottleneck. Follow these guidelines to improve parallelism of the SQL statements: q Because Oracle can parallelize joins more efficiently than subqueries, you should convert your subqueries into joins. q Use PL/SQL functions in the WHERE clause of the main query rather than correlated subqueries. q Queries with distinct aggregates should be rewritten as nested queries. You can create and populate tables in parallel by using the PARALLEL and the NOLOGGING options with the CREATE TABLE statement. For example, CREATE TABLE new_table PARALLEL NOLOGGING AS SELECT col1,col2, col3 FROM old_table; You also can create indexes by using the PARALLEL and NOLOGGING clauses of the CREATE INDEX statement. Index creation takes place serially unless you specify the PARALLEL clause. An index created with an INITIAL of 5MB and a PARALLEL DEGREE of 8 will use at least 40MB during index creation because the STORAGE clause refers to the storage of each subindex created by the query server processes. The number of CPUs can affect the amount of parallelism If the degree of parallelism isn't specified in the PARALLEL clause of CREATE INDEX, the number of CPUs is used as the degree of parallelism. Another technique to optimize parallel operations is to set the initialization parameters correctly: q OPTIMIZER_PERCENT_PARALLEL. The default value of 0 will cause the least usage of resources and will generally result in a long response time. On the other hand, a value of 100 will cause the optimizer to use a parallel plan, unless a serial plan is faster. The recommended value is 100 divided by the number of concurrent users. Parallelism is influenced by the usage of hints A non-zero setting of OPTIMIZER_PERCENT_PARALLEL is overridden if you use a FIRST_ROWS hint or set OPTIMIZER_MODE to FIRST_ROWS.
q

PARALLEL_MAX_SERVERS. The recommended value is 2CPUsNumber of Concurrent Users. PARALLEL_MIN_SERVERS. It's recommended that you set this to the same value as PARALLEL_MAX_SERVERS. SHARED_POOL_SIZE. Oracle reserves memory from the SHARED_POOL for the parallel server processes. You can use the following formula to determine the setting of this parameter: (CPUs + 2)(PARALLEL_MIN_SERVERS) 1.5(BLOCK_SIZE)

Set PARALLEL_MIN_SERVERS appropriately You can use the V$PQ_SYSSTAT view to determine if you've set the value of PARALLEL_MIN_SERVERS too low or too high. If the Servers Started statistics are continuously increasing, you need to increase this parameter. On the other hand, if very few parallel server processes are busy at any given time, you should decrease this value.
q

q q q

SORT_AREA_SIZE. Use a large SORT_AREA_SIZE because the parallel queries would generally be doing some sort of sorting. A small SORT_AREA_SIZE could lead to a lot of sort runs. PARALLEL_ADAPTIVE_MULTI_USER. The recommended value for this is FALSE. When this parameter is set to TRUE, it automatically reduces the requested degree of parallelism based on the current number of active parallel execution users on the system. The effective degree of parallelism is based on the degree of parallelism set by the table attributes or hints, divided by the total number of parallel execution users. This parameter works best for a single-node SMP machine but can be used in an OPS environment if all the following conditions are true: Users that execute parallel operations connect to the same node Each node is an SMP Instance groups aren't configured

http://www.informit.com/content/0789716534/element_022.shtml (9 of 10) [26.05.2000 17:13:25]

informit.com -- Your Brain is Hungry. InformIT - Tuning for Different Types of Applications From: Using Oracle8

< Back

Contents

Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_022.shtml (10 of 10) [26.05.2000 17:13:25]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

You are here : Home : Using Oracle8

Exact Phrase All Words Search Tips

Diagnosing and Correcting Problems

< Back Contents Next > Save to MyInformIT From: Using Oracle8 Author: David Austin Publisher: Que More Information

q q

Common Problems in Oracle Systems Data Block Corruption

r r

Using the DB_VERIFY Utility Using the Checksum Facilities

q q

Recovering from Damaged Data Blocks Working with the Alert Log and Trace Files
r

Oracle Alert Log Oracle Trace Files Other Useful Files Library Cache System Statistics Wait Events Latch Statistics Rollback Segment Contention Shared Pool Size I/O Statistics

Summary errors
r r

Analyze Your Database with UTLBSTAT/UTLESTAT

r r r r r r r

Working with Oracle Support

People encounter four common types of failures or problems in an Oracle system: q Crash (code failure). This can be seen in the form of ORA-600 errors, exception signals (ORA-7445 in UNIX), or other error messages. When a process crashes, it places diagnostic information in a trace file, and the background process places an entry in the alert log to indicate what this trace file is called. Usually, the trace file contains sufficient information to identify the cause of the problem. If the information isn't sufficient, you may need to get additional trace or dump files as requested by Oracle Support Services. From the trace file, the most useful diagnostic information is obtained from the stack trace and process state dump sections. q Hang. You first have to identify the situation as being a hanging situation. If a process is consuming CPU, it isn't hung; a hang is indicated by a process waiting for something that will never happen. The hung process usually isn't the cause of the problem, so you need to obtain more diagnostic information to find the root cause. The most useful tools to diagnose the hanging situation is to get system state dumps from the trace file and v$session_wait. q Looping. A looping situation is identified by a process endlessly repeating the same task and using all available CPU. Loops are difficult to diagnose because the target is moving continuously. You can diagnose this problem by obtaining stack trace and process state dumps information. You'll need to get multiple dumps to identify the location and scope of the loop. q Slow process. A slow system or process is generally the result of insufficient tuning, an art that requires you to

http://www.informit.com/content/0789716534/element_023.shtml (1 of 14) [26.05.2000 17:14:31]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

understand the application and the system environment properly. Oracle provides two scripts, UTLBSTAT and UTLESTAT, which you can use to tune the system by identifying the sources of contention.

Data Block Corruption

An Oracle data block is written in an internal binary format and is considered to be corrupt if it no longer conforms to this format. Oracle checks a block for possible corruption before using it in any way. This checking is performed at the cache layer and higher layers of the Oracle code. The cache layer information includes the block type, version, DBA, incarnation, sequence, and block checksum. If an inconsistency is found, the block is considered to be media corrupt. If a block isn't media corrupt but corruption is identified at the higher layer of the code, the block is considered to be software corrupt. When a block is found to be corrupt, you normally would see an ORA-1578 or ORA-600 [3374] error. You also can use the following methods to detect and identify corrupt or damaged blocks: q The most detailed block check available is the following (this command requires the database to be open): Analyze table table_name validate structure cascade; It checks every table block at the cache and higher levels. It also checks the associated index blocks and checks the one-to-one correspondence between data and index rows. You can use the DB_VERIFY utility to check the validity of Oracle data files, even if the database is down or all the files aren't available.

Using the DB_VERIFY Utility

The DB_VERIFY utility lets you to perform integrity verification of database files. It's a command-line utility that performs a physical data structure integrity check on an offline database; you can find it in the rdbms/admin directory. (DB_VERIFY checks are limited to cache-managed blocks.) Its syntax is as follows: DB_VERIFY parameters Validate your backups You also can use DB_VERIFY to validate your backups and make sure that you can use them for database recovery. In the syntax, DB_VERIFY is an operating system-specific executable: dbv on Sun/Sequent systems and dbverif80 on Windows NT systems. parameters is from the following table: Parameter FILE START END BLOCKSIZE LOGFILE FEEDBACK n HELP PARFILE Description Name of file to verify Starting block address Ending block address Logical block size File that logs the DB_VERIFY's output Displayed for every n pages verified (0 indicates off) Type Y to displayonline help Name of the parameter file None Default None First block of the file Last block of the file 2048 STANDARD ERROR

Specify a BLOCKSIZE If you don't specify BLOCKSIZE for non-2KB files, you'll get error DBV-00103. You can obtain the BLOCK_SIZE by executing the following at the Server Manager prompt: SVRMGR>SHOW PARAMETER BLOCK_SIZE To use the DB_VERIFY utility in Windows NT, enter the following at a system prompt: dbverif80 file=datafile1.ora logfile=dbvlog.out feedback=0 DB_VERIFY will verify the file datafile1.ora, starting from the first Oracle block to the last Oracle block, using a block size of 2048 and put the results in the file dbvlog.out. It doesn't send any dots to the screen for verified pages. Shut down the database before using DB_VERIFY

http://www.informit.com/content/0789716534/element_023.shtml (2 of 14) [26.05.2000 17:14:31]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

You must shut down the database before using DB_VERIFY against its data files, to prevent the data-base from getting corrupted. To use the DB_VERIFY utility on a Sun machine, enter the following at a system prompt: dbv file=datafile2.dbf feedback=10 Validating archive logs You can verify inactive archive log files with DB_VERIFY, even if the database isn't offline. You'll get the following output: DBVERIFY: Release x.x.x.x.x - date Copyright...... DBVERIFY - Verification starting: FILE = datafile2.ora DBVERIFY - Verification complete Total Total Total Total Total Total Total Total Pages Pages Pages Pages Pages Pages Pages Pages Examined.............................: 9216 Processed....(Data)..................: 2044 Failing.........(Data)..................: 0 Processed....(Index)..................: 921 Failing.........(Index).................: 0 Empty................................: 5442 Marked Corrupt..........................: 0 Influx..................................: 0

On other UNIX platforms such as HP and DEC-UX, results should be very similar. Verifying a Data File Created on a Raw Device The following shows the contents of a raw device when used with a data file:

Data Space left in the data file Reserved for non-data file usage When you use DB_VERIFY to verify a data file on a raw device, you should use the START and END parameters. Otherwise, it will mark the non-database blocks as corrupt: $Dbv testfile.dbf DBVERIFY: Release x.x.x.x.x - date Copyright........ DBVERIFY Page 23548 Page 23600 Page 23601 Page 23602 Page 23603 Page 23604 Page 23605 Page 23606 Page 23607 Page 23608 Page 23609 Page 23610 Verification starting: FILE = is marked software corrupt is marked software corrupt is marked software corrupt is marked software corrupt is marked software corrupt is marked software corrupt is marked software corrupt is marked software corrupt is marked software corrupt is marked software corrupt is marked software corrupt is marked software corrupt testfile.dbf

http://www.informit.com/content/0789716534/element_023.shtml (3 of 14) [26.05.2000 17:14:31]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

Page Page Page Page Page Page Page Page Page

23611 23612 23613 23614 23615 23616 23617 23618 23723

is is is is is is is is is

marked marked marked marked marked marked marked marked marked

software software software software software software software software software

corrupt corrupt corrupt corrupt corrupt corrupt corrupt corrupt corrupt

DBVERIFY - Verification Complete Total Total Total Total Total Total Total Total Pages Pages Pages Pages Pages Pages Pages Pages Examined..............................: 12075 Processed....(Data).......................: 0 Failing.........(Data)....................: 0 Processed....(Index)....................: 462 Failing.........(Index)...................: 0 Empty.................................: 11482 Marked Corrupt...........................: 20 Influx....................................: 0

Ensure that you don't get blocks corrupted when using DB_VERIFY 1. At the Server Manager prompt, type the following: SVRMGR> select bytes from v$datafile where name = 'datafile_name'; By using the bytes amount from this query, you can determine the number of database blocks in the data file with the following equation: Number of blocks = datafile bytes/BLOCKSIZE 2. Run DB_VERIFY: $Dbv file=datafile end=number of blocks

Using the Checksum Facilities

The checksum facilities provided by Oracle can be used to identify corrupt blocks. Setting DB_BLOCK_CHECKSUM to TRUE causes checksums to be calculated for all data blocks on their next update. A block will always use checksums as soon as it's generated, even if the parameter is removed. The checksums are calculated by the DBWR and the direct loader and stored in the cache header of every data block when writing it to disk. Setting LOG_BLOCK_CHECKSUM to TRUE causes checksums to be calculated for all redo log blocks. Parameters used to check for corrupt blocks can affect performance Setting DB_BLOCK_CHECKSUM or LOG_BLOCK_CHECKSUM to TRUE can result in CPU time being spent in validating blocks, a process that's normally not performed. The performance hit can be from 5% to 20%, depending on the type of data. Setting _DB_BLOCK_CACHE_PROTECT to TRUE in the init.ora file protects the cache layer from becoming corrupted. It will prevent certain corruption from getting to disk, although it may crash the foreground of the database instance. It will help catch stray writes in the cache and will fail with a stack violation when a process tries to write past the buffer size in the SGA. The checksum utility also provides various events that you can use to check various blocks for corruption. These events "soft corrupt" a block if they find any kind of mismatch during the checking process. To use these events, type the following at the Server Manager prompt: alter session set events '102xx trace name context forever, level 10'; The following events are available: Event 10210 10211 10212 Purpose Checks data blocks Checks index blocks Checks cluster blocks

http://www.informit.com/content/0789716534/element_023.shtml (4 of 14) [26.05.2000 17:14:31]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

10225

Checks information in fet$/uset$ for any corruption; can be used in case your create segment statement is hanging

Recovering from Damaged Data Blocks

There are several causes for a data block getting damaged: q Bad I/O hardware/firmware q Operating system I/O or caching problems q Disk repair utilities q Memory problems q Oracle incorrectly attempting to access an unformatted block To recover from damaged data blocks, you need to figure out the extent of the damage by looking at the full error message(s), the alert log, and all the trace files generated. If there are multiple errors or you have operating system-level errors, you should first resolve those errors before even attempting to recover the damaged file. After the operating system and hardware errors are fixed, you can use the following steps to recover. The error is in file# (F) and block# 1. Determine the corrupted file by typing the following at the Server Manager prompt: SELECT * FROM v$datafile WHERE file# = F; 2. Determine the damaged object by executing the following at the Server Manager prompt: SELECT owner, segment_name, segment_type FROM dba_extents WHERE file_id = F AND B BETWEEN block_id AND block_id + blocks - 1; SEE ALSO If the segment returned from this query is a rollback segment, you can recover from the damaged segment by using a certain method, 3. If the segment is an index, determine the table it belongs to by executing the following: SELECT table_owner, table_name FROM dba_indexes WHERE index_name = 'segment_name'; If the segment is a cluster, determine the table it belongs to by executing the following: SELECT owner, table_name FROM dba_tables WHERE cluster_name = 'segment_name'; If the segment is a table, note the table name and the owner. SEE ALSO If the table is a data dictionary table, recover from the damage by using a described method, 4. Run the ANALYZE command at least twice to ensure that the problem isn't an intermittent problem. For an index or a table, at the Server Manager prompt type the following: ANALYZE TABLE owner.tablename VALIDATE STRUCTURE For a cluster, type the following at the Server Manager prompt: ANALYZE CLUSTER owner.clustername VALIDATE STRUCTURE CASCADE;

CASCADE;

If you encounter potential hardware errors on a particular disk or controller, first relocate the files to a good disk. Recover from a hardware problem in ARCHIVELOG mode 1. Take the affected data file offline. 2. Restore its last backup on a good disk. 3. Rename the data file to the new location. 4. Recover the data file. 5. Put the file back online and start using it. Recover from a hardware problem in NOARCHIVELOG mode
http://www.informit.com/content/0789716534/element_023.shtml (5 of 14) [26.05.2000 17:14:31]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

1. 2. 3. 4.

Take the affected data file offline. Restore its last backup on a good disk. Rename the data file to the new location. Put the file back online and start using it.

If you have to rename the file, run the ANALYZE command at least twice to verify that the corruption still exists. If ANALYZE still returns an error and you've already fixed the hardware problems, you need to salvage the data. You can salvage data from damaged data blocks in several ways: r Media recovery is the easiest method to resolve block corruption problems. r If the object (table, index, whatever) that contains corruption can be easily recreated, you should drop the object and recreate it. If you know the file and block number of the corrupted blocks, you can extract the good data by selecting around the corruption. Before attempting to salvage the data this way, check the following: r Do you have an export that can be used to create the table easily? r Do you have a backup copy of the database from which you can create an export of the table? If you have neither, you can extract data around the corrupt block by using the following command at the Server Manager prompt: (a) CREATE TABLE salvage AS SELECT * FROM corrupt_table WHERE 1 = 2; (b) INSERT INTO salvage SELECT /*+ ROWID(corrupt_table) */ * FROM corrupt_table WHERE rowid <= 'low_rowid_of_corrupt_block'; (c) INSERT INTO salvage SELECT /*+ ROWID(corrupt_table) */ * FROM corrupt_table WHERE rowid >= 'high_rowid_of_corrupt_block'; You also can set event 10231 in the initialization file and select around the corruption. This event causes Oracle to skip software and media-corrupted blocks when performing full table scans and allows you to extract the good blocks and recreate the database object: Usage: event="10231 trace name context forever, level 10" You can set event 10233 in the initialization file and select around the corruption. This event is similar to event 10231, except it works with index range scans. Note, however, that data in the corrupted blocks will be lost when event 10233 is set. Finally, you can contact Oracle Support Services, which has access to several tools that you can use to extract data from corrupt database objects. See the later section "Working with Oracle Support" for more information.

Working with the Alert Log and Trace Files

In an Oracle environment, you'll encounter many kinds of logs and trace files, including the alert log, process trace files, core files, application logs, system logs, and network logs.

Oracle Alert Log

The Oracle background and foreground processes log information to a file called the alert log. The primary use of the alert log is to understand the progress of the instance, because it captures the following:

Summary errors
r r

The important stages the database goes through, such as startup, shutdown, and tablespace creation/drop Pointers to trace files when there's a failure SEE ALSO More information on the contents and usage of the alert log,

Oracle Trace Files

All the Oracle server processes have a file called a trace file to which they record traces, dumps, and error messages. This file isn't written to by any other process. Locating trace files

http://www.informit.com/content/0789716534/element_023.shtml (6 of 14) [26.05.2000 17:14:31]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

You should find the trace file in the directory indicated by BACKGROUND_DUMP_DEST,USER_DUMP_DEST,or CORE_DUMP_DEST, depending on the exact error and its cause. The trace file is linked to the standard output of the process, allowing the capture of encountered operating system messages. All trace files have a header that contains r Timestamp r Oracle version used r Operating system and version r Installed options r Instance name r Oracle process ID r Operating system process ID Application Trace Files Client-side progress and failure information can be collected in application trace files-for example, spool.lst from SQL*Plus. These files contain useful information, such as the following: r Application progress and performance r Alert of application malfunction r Indications of system, network, or database problems Extracting the Stack Trace Out of the Core Dump When a process aborts, it creates a core file in the current directory. This core file contains a dump of process memory. To dump your process state, use the following command: Alter session set events 'immediate trace name processstate level 10'; You can extract a stack trace from the core file, which can indicate where the process failed. To obtain a stack trace when a certain error XXXX occurs, use the following command: Alter session set events 'XXXX trace name errorstack forever, level 10'; UNIX versus Windows NT core dump These steps are UNIX specific. If you use Windows NT, you can find the core dump in an "access violation" file. Get the stack trace from the core file 1. Log in as ORACLE and change to the $ORACLE_HOME/bin directory. 2. Type the following, where program is the program that aborted: file program 3. Add read permissions to the program. At the operating system prompt, type $Chmod +r program 4. Log out and then log in as the user who encountered the error. 5. The next step varies, depending on the version of UNIX you're using. One of the following commands should exist on your machine: Command dbx xdb sdb adb gdb Exit Command/Keystroke quit quit q Ctrl+D Ctrl+D Change to the directory where the core dump is located. In the Bourne or Korn shell, type the following: View Code In the C shell, type the following:

http://www.informit.com/content/0789716534/element_023.shtml (7 of 14) [26.05.2000 17:14:31]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

dbx $ORACLE_HOME/bin/program core | tee /tmp/stacktrace 6. The stack trace should be produced in the file stacktrace. Exit the debug tool.

Other Useful Files

Several other useful files can help you diagnose system problems: r You can use system logs to obtain information about hardware or operating system problems. r Use network logs to find network messages.

Analyze Your Database with UTLBSTAT/UTLESTAT

Oracle provides you with two scripts, utlbstat.sql and utlestat.sql, that you can use to tune a database. The utlbstat component gathers the initial performance statistics and places them in temporary tables; the utlestat component gathers the performance statistics at the end of the observation period and places them in temporary tables. When utlestat is finished, it compares the information in the two sets of temporary tables and places the result of comparison in another set of temporary tables. A report is generated from the results and is placed in an ASCII report.txt file in the current directory. UTLBSTAT and UTLESTAT don't affect performance Using UTLBSTAT and UTLESTAT doesn't result in a performance degradation, except for a few minutes while these scripts run. This is because these scripts simply take a snapshot of the system at the beginning and a snap-shot at the end of the period. Use the utlbstat and utlestat scripts 1. Determine the time period for analysis. You should choose the time period to represent your normal workload. You shouldn't run utlbstat right after the database is started, for example, because it won't represent the normal working condition of the database. Let the database run for some time so that the system caches can be loaded appropriately and represent a stable running environment. 2. Set initialization parameters. In the init.ora file, set TIMED_STATISTICS to TRUE and then restart the database. Alternatively, if you don't want to restart the database, connect internally and type the following at the Server Manager prompt: ALTER SYSTEM SET TIMED_STATISTICS = TRUE; 3. Run utlbstat.sql from the Server Manager prompt at the appropriate time: SVRMGR> @%RDBMS80%\admin\utlbstat 4. Run utlestat.sql from the Server Manager at the end of the time period: SVRMGR>@%RDBMS80%\admin\utlestat Keep the database running while getting reports Don't shut down the database between the running of UTLBSTAT and UTLESTAT; otherwise, the results will be useless. The resulting report.txt file provides the following information: r Library cache statistics r Systemwide statistics r Wait events r DBWR statistics r I/O statistics r Latch statistics r Dictionary cache statistics Let's analyze each section in more detail and determine how to tune the system based on the report's information.

Library Cache
The SQL AREA, TABLE/PROCEDURE, BODY, and TRIGGER rows in the following output show library cache activity for SQL statements and PL/SQL blocks. The other rows indicate library cache activity for object definitions used by Oracle for dependency management.

http://www.informit.com/content/0789716534/element_023.shtml (8 of 14) [26.05.2000 17:14:31]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

How many times object was found in memory (equivalent to # of parses) How many times did object was executed Object definition was aged out for lack of space Should be greater than 0.9 for all rows Should be (ideally 0) not more than 1 percent of PINS Your aim is to reduce parsing, enable sharing of statements, reduce aging out, and provide enough space to large objects. If you have a lot of reloads or the GETHITRATIO is less than 90%, you should increase the SHARED_POOL_SIZE parameter in the initialization file.

System Statistics
The following file also provides several comments (not shown here) that you can read for further information: Improving database writerperformance From these system statistics, DBWR buffers scanned row and DBWR checkpoints should give you a good idea about the amount of load on the DBWR process. On operating systems such as UNIX that allow you to have multiple DBWR processes, you should increase the parameter DB_WRITERS in the initialization file to two per database file. Also, increase DB_BLOCK_WRITE_BATCH to reduce the number of times the DBWR is signaled to perform a write operation. Statistic Total Per Transa Per Logon Per Sec -------------------------- ----- ---------- --------- ------CPU used by this session 9904 8.94 70.24 27.66 CPU used when call started 9904 8.94 70.24 27.66 CR blocks created 9 .01 .06 .03 DBWR buffers scanned 907 .82 6.43 2.53 DBWR checkpoints 6 .01 .04 .02 DBWR free buffers found 763 .69 5.41 2.13 DBWR lru scans 89 .08 .63 .25 DBWR make free requests 76 .07 .54 .21 DBWR summed scan depth 907 .82 6.43 2.53 DBWR timeouts 91 .08 .65 .25 OS System time used 284700 256.95 2019.15 795.25 OS User time used 838900 757.13 5949.65 2343.3 SQL*Net roundtrips to/from 8899 8.03 63.11 24.86 background checkpoints comp 7 .01 .05 .02 background checkpoints star 6 .01 .04 .02 background timeouts 214 .19 1.52 .6 bytes received via SQL*Net1167468 1053.67 8279.91 3261.08 bytes sent via SQL*Net to c343632 310.14 2437.11 959.87 calls to get snapshot scn: 5622 5.07 39.87 15.7 calls to kcmgas 1130 1.02 8.01 3.16 calls to kcmgcs 101 .09 .72 .28 calls to kcmgrs 9064 8.18 64.28 25.32 change write time 529 .48 3.75 1.48 cleanouts only - consistent 1 0 .01 0 cluster key scan block gets 334 .3 2.37 .93 cluster key scans 303 .27 2.15 .85 commit cleanout failures: b 1 0 .01 0 commit cleanout number succ 1329 1.2 9.43 3.71 consistent changes 9 .01 .06 .03 consistent gets 5784 5.22 41.02 16.16 cursor authentications 3744 3.38 26.55 10.46 data blocks consistent read 9 .01 .06 .03 db block changes 18659 16.84 132.33 52.12 db block gets 15638 14.11 110.91 43.68

http://www.informit.com/content/0789716534/element_023.shtml (9 of 14) [26.05.2000 17:14:32]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

deferred (CURRENT) block cl 1260 1.14 8.94 3.52 enqueue conversions 139 .13 .99 .39 enqueue releases 2472 2.23 17.53 6.91 enqueue requests 2466 2.23 17.49 6.89 execute count 10859 9.8 77.01 30.33 free buffer requested 447 .4 3.17 1.25 immediate (CR) block cleano 1 0 .01 0 logons cumulative 141 .13 1 .39 logons current 1 0 .01 0 messages received 684 .62 4.85 1.91 messages sent 684 .62 4.85 1.91 no work - consistent read g 3291 2.97 23.34 9.19 opened cursors cumulative 3655 3.3 25.92 10.21 opened cursors current 3 0 .02 .01 parse count 5728 5.17 40.62 16 parse time cpu 1602 1.45 11.36 4.47 parse time elapsed 1799 1.62 12.76 5.03 physical reads 29 .03 .21 .08 physical writes 1021 .92 7.24 2.85 recursive calls 25096 22.65 177.99 70.1 recursive cpu usage 5052 4.56 35.83 14.11 redo blocks written 1420 1.28 10.07 3.97 redo buffer allocation retr 11 .01 .08 .03 redo entries 9339 8.43 66.23 26.09 redo log space requests 13 .01 .09 .04 redo log space wait time 856 .77 6.07 2.39 redo size 1796924 1621.77 12744.14 5019.34 redo small copies 1359 1.23 9.64 3.8 redo synch time 5011 4.52 35.54 14 redo synch writes 565 .51 4.01 1.58 redo wastage 1076955 971.98 7637.98 3008.25 redo write time 5529 4.99 39.21 15.44 redo writer latching time 7 .01 .05 .02 redo writes 994 .9 7.05 2.78 rollback changes - undo rec 278 .25 1.97 .78 rollbacks only - consistent 9 .01 .06 .03 session logical reads 21135 19.07 149.89 59.04 session pga memory 20645272 18632.92 146420.3757668.36 session pga memory max 20645272 18632.92 146420.3757668.36 session uga memory 232400 209.75 1648.23 649.16 session uga memory max 5826432 5258.51 41322.2116274.95 sorts(disk) sorts (memory) 282 .25 2 .79 sorts (rows) 3414 3.08 24.21 9.54 table fetch by rowid 554 .5 3.93 1.55 table fetch continued row indicates row chaining table scan bl locks gotten 571 .52 4.05 1.59 table scan rows gotten 3207 2.89 22.74 8.96 table scans (long tables) 1 0 .01 0 table scans (short tables) 833 .75 5.91 2.33 total number commit cleanou 1330 1.2 9.43 3.72 user calls 7271 6.56 51.57 20.31 user commits 1108 1 7.86 3.09 user rollbacks # of rollback calls issued by users. User rollbacks/user commits, if high indicates a problem. write requests 109 .1 .77 .3 Use the following formulas to calculate the data cache hit ratio: LOGICAL READS = CONSISTENT GETS + DB BLOCK GETS HIT RATIO = (LOGICAL READS - PHYSICAL READS) / LOGICAL READS By using these calculations for the preceding output, we get LOGICAL READS = 5784 + 15638 = 21422

http://www.informit.com/content/0789716534/element_023.shtml (10 of 14) [26.05.2000 17:14:32]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

CACHE HIT RATIO = (21422 - 29) / 21422 = 99.86% If the cache hit ratio as calculated here is less than 80%, you should increase the DB_BLOCK_BUFFERS parameter in the initialization file. Use the following equation to check the effectiveness of your application to see if indexes are used properly. The result should be close to zero. Non-Index lookups ratio = table scans (long tables) / table scans (long tables) + table scans (short tables) = 1/(1+833) = close to zero.

Wait Events
Event Name Count Total Time Avg Time ----------------------------- ----- ---------- -------SQL*Net message from client 9385 41677 4.44 log file sync 563 5196 9.23 write complete waits 68 1624 23.88 log file switch completion 13 856 65.85 log buffer space 7 301 43 SQL*Net message to client 9385 54 .01 buffer busy waits 10 25 2.5 db file sequential read 36 18 .5 SQL*Net more data from client 139 12 .09 latch free 4 2 .5 control file sequential read 22 0 0 db file scattered read 1 0 0 Your goal is to eliminate all waits for resources: Buffer Busy Waits ratio = Buffer busy waits / Logical reads = 11 / 21422 = close to zero Minimize the waits for buffers A ratio of buffer busy waits greater than 4 percent indicates that you need to tune the DB_BLOCK_BUFFERS parameter. For example, r If the waits are due to data blocks, you should increase the FREELISTS parameter for the heavily inserted table. r If you see waits for UNDO segments, you should add more rollback segments. r If the Sorts (memory) are less than 90% of the total sorts, you should increase the SORT_AREA_SIZE parameter in the initialization file. r If you see a lot of enqueue waits, you should increase the parameter ENQUEUE_RESOURCES in the initialization file.

Latch Statistics

LATCH_NAME GETS MISSES HIT_RATIO SLEEPS SLEEPS/MISS ------------------ ------ ------ --------- ------ ----------cache buffers chai 70230 2 1 0 0 cache buffers lru 1540 0 1 0 0 dml lock allocatio 2545 0 1 0 0 enqueue hash chain 5051 0 1 0 0 enqueues 6574 0 1 0 0 ktm global data 1 0 1 0 0 latch wait list 6 0 1 0 0 library cache 136074 19 1 3 .158 library cache load 2 0 1 0 0 list of block allo 2234 0 1 0 0

http://www.informit.com/content/0789716534/element_023.shtml (11 of 14) [26.05.2000 17:14:32]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

messages modify parameter v multiblock read ob process allocation redo allocation row cache objects sequence cache session allocation session idle bit session switching shared pool sort extent pool system commit numb transaction alloca undo global data user lock 26 rows selected.

4122 140 6 139 12711 15913 431 3361 15861 6 10198 1 16081 3356 3440 556

2 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Anything less than 0.98 indicates a potential problem SEE ALSO More information on reducing latch contention,

Rollback Segment Contention

If the waits-to-gets ratio is greater than 5%, consider adding rollback segments Used in determining the waits-to-gets ratio Shouldn't be high; set OPTIMAL accordingly

Shared Pool Size

The following file also includes columns for COUNT and CUR_USAG (not shown here): NAME GET_REQS GET_MISS SCAN_REQ SCAN_MIS MOD_REQS -------------- -------- -------- -------- -------- -------dc_tablespaces 283 0 0 0 0
http://www.informit.com/content/0789716534/element_023.shtml (12 of 14) [26.05.2000 17:14:32]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8

dc_free_extent dc_segments dc_rollback_se dc_used_extents dc_tablespace_q dc_users dc_user_grants dc_objects dc_tables dc_columns dc_table_grants dc_indexes dc_constraint_d dc_constraint_d dc_usernames dc_sequences dc_tablespaces 18 rows selected.

15 3 12 3 3 1272 849 166 2724 596 1260 2 1 0 282 7 3

3 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0

3 0 0 0 0 0 0 0 0 56 0 16 13 1 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

9 3 0 3 3 0 0 0 0 0 0 0 0 0 0 7 3

If the GET_MISS to GET_REQS ratio is greater than 15%, consider increasing SHARED_POOL_SIZE.

I/O Statistics
The following file also includes READ_TIME and MEGABYTES columns (not shown). The MEGABYTES column shows the size of the tablespaces.

Try to balance this out overall the tablespaces The following file also includes WRITE_TIME, READ_TIME, and MEGABYTES columns (not shown). TABLE_SPACE FILE_NAME READS BLKS_READ WRITES BLKS_WRT ----------- ----------- ----- --------- ------ -------RBS rbs01.dbf 1 1 663 663 SYSTEM system01.d 38 46 372 372 TEMP temp01.dbf 0 0 0 0 TOOLS tools01.dbf 0 0 0 0 USERS users01.dbf 0 0 0 0 5 rows selected. The following actions can be taken to balance out the load: r Move one or more database files to another disk. r Frequently accessed tables should be separated from other tables and moved to their own data files on another drive. r Separate your rollback segments, redo logs, and archive logs.

Working with Oracle Support

You can do several things to obtain quality service from Oracle Support. These things are very simple but important because they tell Oracle Support the exact nature and urgency of your problem, so they can provide you with the best possible resolution:

http://www.informit.com/content/0789716534/element_023.shtml (13 of 14) [26.05.2000 17:14:32]

informit.com -- Your Brain is Hungry. InformIT - Diagnosing and Correcting Problems From: Using Oracle8
r

Keep your Customer Support information handy. Most often, problems occur when you don't want them to happen. Not having contact information readily available can result in loss of valuable time. Keep the contact telephone numbers and your Customer Support Identification (CSI) number around. Also make sure that you understand the type of customer service contract you have with Oracle Support Services. Understand the problem. This will allow your problem to be addressed by the correct group at Oracle Support Services. Set the correct priority for your problem. Make sure that the analyst who obtains information from you is made aware how critical your problem is. Provide configuration information. When contacting Oracle Support Services, you should have as much of the following information as possible: s The hardware and operating system release number on which the application(s) is running s The release number of all Oracle products involved in the problem s A clear problem description s Any third-party vendor and version in use s The error messages s The alert log s The trace files that you've obtained in relation to the error Provide test cases when requested. Several times the problem isn't easy to describe, and it helps to provide a test case to Oracle Support. A test case is a collection of trace files, code modules, and steps that can be used to reproduce the problem. You can assume that the analyst at Oracle Support doesn't know much about the application specifics, so you should try to create a very simple test case. If you can, try to use the standard Oracle schema (such as SCOTT) and standard demo tables (such as EMP and DEPT) for your test case. Provide the steps to reproduce the problem and keep the size of the test case minimal. Have your Technical Assistance Request (TAR) number handy. Oracle Support Services assigns a TAR number (a PMS number in some countries) to every problem you report. If you have multiple issues, a new TAR will be generated for each issue. Every time you contact Oracle Support, you should refer to this TAR number. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_023.shtml (14 of 14) [26.05.2000 17:14:32]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

You are here : Home : Using Oracle8

Configuring and Using Net8 Features

Exact Phrase All Words Search Tips

< Back

Contents

Next >

Save to MyInformIT

From: Using Oracle8 Author: David Austin Publisher: Que More Information

Introducing Net8
r r

Net8 Naming Techniques Supported Network Protocols TNSNAMES.ORA LISTENER.ORA SQLNET.ORA SUPPORT (value is 16) Multiplexing Dead Connection Detection Parallel Server Reconnections The Default TNSNAMES.ORA File

Basic Net8 Files

r r r r

Net8 Features
r r r

Configuring Net8
r

q q

Connecting to Other Products with ODBC Connecting to Multithreaded Servers

r r r r

Setting Net8 Entries and Initialization Parameters Connecting to Shared Servers Managing Dispatchers Managing Shared Servers

Net8 is an Oracle networking product that facilitates network communication between remote clients and database servers, as well as between two or more database servers. Net8, the successor to SQL*Net 2.x, is a software layer that runs on top of standard network protocols such as TCP/IP and SPX/IPX. Like its predecessor, Net8 enables location, network, and application transparency: q Network transparency enables Net8 and client-side tools to work the same, regardless of network protocols between server and clients. Oracle network protocol adapters that work identically across different network protocols facilitate network transparency. You can change the network structure without making any changes at the application level. q Location transparency is the capability of making remote database objects appear local. This is done with the use of synonyms and links. If an instance is moved from its location, changes are transparent to the application. q Application transparency gives transparency to database objects when they're migrated between systems. An application tested against a test instance and being deployed on a production instance is an example.

http://www.informit.com/content/0789716534/element_024.shtml (1 of 13) [26.05.2000 17:15:35]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

Net8 Naming Techniques

You can implement Net8 by using host, local, external naming, and Oracle Names server methodologies. Oracle Names is a centralized management system used mostly by larger shops. An Oracle Names Server maintains client configuration information in one location. Oracle Names Server is very similar to Domain Name Service (DNS) on IP-based networks. Each method has its advantages and disadvantages. Host, local, and external naming are most commonly used for smaller installations. It really doesn't matter what naming convention and technique a shop uses. What really matters is that the technique is unique. Whenever connections are made between clients and servers, there must be no confusion as to what database the connection must be made to. To enable uniqueness, Oracle and Net8 use the convention of DB_NAME.DB_DOMAIN. A domain is a logical collection of machines, hardware, and networked devices in a networked environment. The convention is that, all names must be unique within a domain, but can be repeated across domains. If your shop uses DNS (Domain Naming Service) for defining its network, you can use DNS-type conventions for your Oracle domain. For example, if your server's host name in a DNS architecture is ABCINC.STARSHIP.COM, you could keep the domain name ABCINC.STARSHIP.COM, but it's not necessary to do so. You may want to keep the domain name WORLD for backward compatibility, especially if you have a heterogeneous environment (with several networking operating systems at work). However, if your database server is serving intranets and the Internet, observing DNS naming conventions may be more advantageous to your shop. The DB_DOMAIN parameter in INITSID.ORA acts as an extension component to a global database name. WORLD is DB_DOMAIN's default value. The global name is a unique name that represents the database's name and domain. The representation for the database's global name is DB_NAME.DB_DOMAIN. Don't confuse this with the parameter GLOBAL_NAMES in the INITSID.ORA file. GLOBAL_NAMES has to be set to TRUE if you want database links to use the same name as the DB_NAME. As seen in Figure 24.1, if you want to refer to table TOASTERS in the MONTHLY schema, the global name that would prefix MONTHLY.TOASTERS would be SALES or SALES.WORLD because there's only one domain. In contrast, Figure 24.2 shows several domains. If you want to refer to table MONTHLY.TOASTERS in the SAMERICA domain, you would use SALES.SAMERICA.ABCINC as the global name (that is, refer to the database SALES in the domain SAMERICA.ABCINC). Similarly, to access the table MONTHLY.TOASTERS in the EUROPE domain, you would use SALES.EUROPE.ABCINC as the global name. This ensures that uniqueness is maintained when you have similar tables and objects in different locations. Figure 24.1 : This naming structure uses a single domain name. Figure 24.2 : This naming structure uses multiple domain names.

Supported Network Protocols

Net8 works with several industry-standard network protocols. These include but are not restricted to TCP/IP, SPX/IPX, Named Pipes, LU6.2, DECnet, Bequeath, and AppleTalk. The adapters for each protocol are supplied mostly by the vendors of the operating system on which they run. To configure Oracle networking, you need to know the following: q Your shop's network protocol q The Oracle instance name (SID) you want to connect to q The host name or IP address (if you were using TCP/IP) of the server on which the Oracle instance you want to connect to is running q Oracle names if Oracle Names Server is being used Testing network connectivity Because Net8 is a layer on top of an operating system network protocol, it makes sense to ensure that the clients and servers can "see" each other. On IP-based networks, just pingthe server or the client from the server to verify this. If you can't ping the destination host, Net8 won't work and you'll need to contact your network/systems administrator in such a situation. An Oracle utility named TNSPING80 verifies that the listener is listening on the server. TNSPING80 is verified normally after the operating system ping is verified.

http://www.informit.com/content/0789716534/element_024.shtml (2 of 13) [26.05.2000 17:15:35]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

Basic Net8 Files

Net8 uses a minimum of three basic files to establish communications between remote clients and server: TNSNAMES.ORA, LISTENER.ORA, and SQLNET.ORA. These files are identical in form, function, and location on all platforms of Oracle. (You may have SNMP.ORA on some platforms and network protocols as well.) Although you can modify these files by using any text editor, Oracle provides (and recommends) tools to configure these files: Net8 Assistant and Net8 Easy Config. We look at these tools later in the chapter.

TNSNAMES.ORA
This configuration file is used by clients to connect to servers (actually, to a listener service running on the server). Servers also use this file to connect to other servers as clients. A sample TNSNAMES.ORA is created on client software in the ORACLE_HOME/NET80/ADMIN folder. You have to add this file to this folder for server-to-server communications because you will unlikely ever do the standard client installation on the server. Definition of ORACLE_HOME varies The definition of ORACLE_HOME varies on different platforms and depends on the installation. On Windows NT systems, it may be in the form of C:\ORANT. On UNIX servers, an environment variable probably defines ORACLE_HOME. The folder and file structure is virtually identical on all plat-forms under ORACLE_HOME. The following is an example of this file. Your systems administrator is a good source of information about the listener's port number on the server. Different platform? Different SID The naming convention for the system identifier (SID) varies on different platforms. The SID can be up to four alphanumeric characters on Windows NT. The SID can be up to eight alphanumeric characters on most UNIX platforms. Consult your platform-specific documentation for valid SIDs. FinProd.world = (DESCRIPTION= (ADDRESS_LIST= (ADDRESS= (PROTOCOL=TCP) (Host=STARSHIP) (Port=1521) ) ) (CONNECT_DATA=(SID=finance) (source_route=yes) ) ) As you can see, the TNSNAMES.ORA file has two main components: the service name and the address/connect descriptor. In the file, FinProd is an alias or service name that represents a connection to the database instance Finance, running on a server/host named STARSHIP and establishing a connection on port 1521; that is the port on which the listener is listening. (The LISTENER.ORA file on host STARSHIP, running an instance called Finance, will have an entry for port 1521.) In general, the connect descriptor contains information specific to the database server, database instance, and network protocol. Using Oracle tools to configure networking Oracle recommends using Net8 Easy Config and Net8 Assistant to configure your network connections on the client side. All changes and new entries made via the Net8 Easy Config and Net8 Assistant tools are reflected in the TNSNAMES.ORA file.

LISTENER.ORA
This file is located on the server and acts as a configuration file for the listener service/process. This configuration file runs on the server and "listens" to incoming requests. This file has three major components:
http://www.informit.com/content/0789716534/element_024.shtml (3 of 13) [26.05.2000 17:15:35]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

address list, Oracle SID, and listener parameters. When the Oracle product is installed on a database server, the installer gets basic information about the server's network configuration (host name, IP address, or both) and creates a sample LISTENER.ORA in the ORACLE_HOME/NET80/ADMIN folder. If you want more than one listener service on a server, simply add details about the listeners in the listener's ADDRESS_LIST component. If you have one listener service for several Oracle instances (SIDs) on a server, simply add the instance information in this file's SID_LIST component. You can have multiple listeners for a single database listening on different server ports, or you can have a single listener listening on a port for requests made to different databases on the same server. LISTENER.ORA defines the listener's address descriptor besides the instance's (SID) name or the global name of the database instances for which the listener is listening. The listener parameter portion is used for tracing and logging the listener process. The following is a LISTENER.ORA file:

http://www.informit.com/content/0789716534/element_024.shtml (4 of 13) [26.05.2000 17:15:35]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

Internal network protocol (always needed) Service name associated with IPC Domain; follows Internet naming conventions Same as descriptors for TNSNAMES.ORA Number of seconds listener sleeps before responding to LSNRCTL80 STATUS command Time (in seconds) after which listener request is timed out Amount of tracing desired, from 0 to 16 (0 = no tracing, 16 = support-level tracing) For other configuration parameters that go into the LISTENER.ORA file, refer to Appendix B of the Net8 Administrator's Guide in the Oracle documentation set. Checking Listener Status You can verify the status of the listener process by typing LSNRCTL80 STATUS at the server's command line for UNIX and Windows NT servers. (The LSNRCTL80 executable was formerly LSNRCTL with Oracle 7.x and SQL*Net 2.x.) The command results in the following output: Starting and stopping the listener process on different platforms LSNRTCL80 is the executable used to verify and start/stop the listener process. On UNIX servers, you can start and stop the process with the LSNRCTL80 start and LSNRCTL80 stop commands, respectively. To verify that the process is running, use ps -ef on the server. (You must have the required privileges.) For Windows NT servers, use the Services dialog box, which you access through Control Panel. You can see whether the listener is running and set the listener process's startup to automatic or manual. LSNRCTL80 for 32-bit Windows: Version 8.0.3.0.0 - Production on 29-MAR-98 00:43:48 (c) Copyright 1997 Oracle Corporation. All rights reserved.

Connecting to (ADDRESS=(PROTOCOL=IPC)(KEY=finance.world)) STATUS of the LISTENER -----------------------Alias LISTENER Version TNSLSNR80 for 32-bit Windows: Version 8.0.3.0.0 - Production Start Date 24-MAR-98 21:28:09 Uptime 2 days 3 hr. 15 min. 38 sec Trace Level off Security ON SNMP OFF Listener Parameter File C:\ORANT\NET80\admin\listener.ora Listener Log File C:\ORANT\NET80\log\listener.log Services Summary... FINANCE has 1 service handler(s) The command completed successfully

SQLNET.ORA
This configuration file, used by clients and servers, contains information about Oracle Names (if used) and information about other client parameters such as diagnostics, naming conventions, and security. This file is installed automatically on the server in the ORACLE_HOME/NET80/ADMIN folder. Net8 client installations also install it in the same folder on the client side. A sample client-side SQLNET.ORA file is as follows: Distributing the files

http://www.informit.com/content/0789716534/element_024.shtml (5 of 13) [26.05.2000 17:15:35]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

In a client/server environment, you may have several hundreds of personal computers that need to be connected to the Oracle database with Net8. You could use Microsoft Systems Management Server (SMS) to distribute the TNSNAMES.ORA and SQLNET.ORA files to each PC. If you want to make a server a client to another server, you can copy or FTP SQLNET.ORA and TNSNAMES.ORA to the database server. TRACE_LEVEL_CLIENT = OFF sqlnet.authentication_services = (NTS) names.directory_path = (TNSNAMES, HOSTNAME) names.default_domain = world name.default_zone = world automatic_ipc = off The settings for TRACE_LEVEL_CLIENT are similar to the TRACE_LEVEL_LISTENER parameter in LISTENER.ORA. The values for both parameters can be either a scale from 0 through 16 or the following predefined names (which correspond to numeric values): q OFF (value is 0) q ADMIN (value is 6) q USER (value is 4) q SUPPORT (value is 16) AUTOMATIC_IPC can be turned on or off, depending on whether IPC is wanted. A new parameter introduced in Net8, SQLNET.EXPIRE_TIME, is used for dead-connection detection. The recommended value for this is 10 (minutes). This parameter must be entered in the server's SQLNET.ORA file. It's a good idea to consult your network administrator before enabling this parameter; a packet-albeit small-would be sent out at the interval specified in the SQLNET.EXPIRE_TIME parameter.

Net8 Features
The following sections cover a few new Net8 features that have improved the product over SQL*Net 2.x.

Multiplexing
Oracle8 Enterprise Edition provides the Connection Manager, which facilitates multiplexing. Net8 takes in multiple client connections and combines or multiplexes them over a single transport connection through Oracle Connection Manager to the destination database. Multiplexing improves response time and increases the number of client connections. It also better uses resources on the server. Multiplexing is available only on TCP/IP networks and only if the multithreaded server (MTS) option is used. You examine MTS later in the "Connecting to Multithreaded Servers" section. The Connection Manager uses the CMAN.ORA file to configure multiplexing.

Dead Connection Detection

This new feature identifies and terminates "dead" connections caused by improper client session termination. If dead-connection detection is enabled, a probe packet is sent to the client from the server. The client process is terminated if no response is received within a user-defined interval. Dead-connection detection makes administration of the database a little bit easier; uncommitted transactions are rolled back and the locks associated with the transaction are released. In addition, resources associated with connection of the "dead" process aren't wasted.

Parallel Server Reconnections

Network configuration files for Oracle Parallel Server (OPS) have to be configured a little differently. In the TNSNAMES.ORA file, the ADDRESS_LIST should include the details of the nodes (host name, port number) and Oracle instance names (SIDs). This enables a client-side application to go through the list and connect to the first available instance for a service name. If an instance in the ADDRESS_LIST is unavailable, the client application simply goes on to the next available instance.

http://www.informit.com/content/0789716534/element_024.shtml (6 of 13) [26.05.2000 17:15:35]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

Configuring Net8
The following sections review the predefined configurations of Net8 that are part of the default installation of most Oracle client-side software, such as SQL*Plus 8.0. You also see how to use two tools specifically for client-side network configuration: Net8 Assistant and Net8 Easy Config, both of which are written in the Java programming language.

The Default TNSNAMES.ORA File

The sample TNSNAMES.ORA file created during a standard Oracle client installation has default samples that correspond to different network protocols (such as TCP, IPX, and Bequeath). You can ask your network administrator on what port the listener is listening (the default is normally 1521), as well as other details, such as database instance name (SID). Then you can drop, modify, or add aliases (services) to configure the client with the Net8 Assistant and Net8 Easy Config. This is covered in detail later in this chapter's "Using Net8 Assistant" and "Using Net8 Easy Config." Use Net8 Easy Config and Net8 Assistant instead of an editor Although you can modify TNSNAMES.ORA via a text editor such as Notepad, it's highly recommended that you use Net8 Easy Config or Net8 Assistant to modify the client side. Using the tools ensures that only the required changes are made to the existing file. When you use Windows Notepad or a UNIX editor such as vi, you might miss a parenthesis or type some extra, unwanted characters. Oracle8 Server is packaged with a CD-ROM called Oracle8 Client, which creates a standard client configuration, including TNSNAMES.ORA and SQLNET.ORA files on the client. Because your client machine (PC or server) will be connected to the database server with just one networking protocol, you can choose your network protocol during the Oracle8 client installation. If one isn't chosen, the standard client installation creates a TNSNAMES.ORA with the most commonly used network protocols listed. You look at three sections of a default TNSNAMES.ORA file that explain TCP/IP, IPX/SPX, and Extproc:

http://www.informit.com/content/0789716534/element_024.shtml (7 of 13) [26.05.2000 17:15:35]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

TCP section Server's host name SPX section Section for calling external procedures As you can see in the default file, Oracle client-side installation gives you various choices for configuring the TNS alias or service name. The word world represents the domain; the letters prefixing the domain are the service names or aliases. A description of each alias follows each alias.world combination. This description varies for different network protocols. Fortunately, you don't have to remember the syntax for each protocol adapter. It's always useful to know the syntax, though, so that if you have to move the files to a server, you can do so with minimal editing. Oracle8 and Net8 have "external" stored procedure capability; the last service name (extproc_connection_data.world) describes the address for the external procedure. When stored or internal procedures written in PL/SQL programs call external procedures, the Net8 listener spawns a session-specific process and, using PL/SQL, passes information regarding the external procedure name, shared library name, and arguments (if needed). Examples of external programs are shared libraries written in a language such as C++/C. For more information on external procedures, refer to Oracle8 Server Administrator's Guide and PL/SQL User's Guide and Reference in the Oracle documentation set. Using Net8 Assistant Net8 Assistant, a new tool written in the Java programming language, has a similar look and feel on different platforms. It replaces Oracle Network Manager, which was packaged with Oracle Server through release 7.3.4. As Figure 24.3 shows, Net8 Assistant provides a graphical user interface that helps you administer and configure profiles (groupings of Net8 parameters), service names, and Oracle Names Servers. To launch the Net8 Assistant in Windows NT or Windows 95, choose Start, and then Programs, and then click Oracle for Windows NT/9x; select the Oracle Net8 Assistant icon. Figure 24.3 : Create a profile by using Net8 Assistant's general options. Configuring the Profile As you can see in Figure 24.3, the Profile branch of the network tree on the left displays a drop-down box and tabbed pages on the right. Each item in the drop-down box presents different configuration pages. If you select the General option, you can customize Tracing, Logging, Routing, and Advanced settings: q The Tracing page gives you more detailed information about Net8 than the contents of a log file provides. You can activate tracing at the server and client level. You can specify the tracing level by choosing from OFF, USER, ADMIN, and SUPPORT. You also can specify trace directories and filenames for the clientand server-side files. q The Logging page lets you specify log directories and filenames on the server and client. Changing the values of the LOG_DIRECTORY_CLIENT, LOG_DIRECTORY_SERVER, LOG_FILE_CLIENT, and LOG_FILE_SERVER parameters sets the corresponding values in the SQLNET.ORA file.
http://www.informit.com/content/0789716534/element_024.shtml (8 of 13) [26.05.2000 17:15:35]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8
q

With the Routing page, you can route connect requests to specific processes. You can configure Dedicated Server, Interprocess Communication (IPC), Addressing for Client, and Source Route Addresses. If Dedicated Server is set to ON, the listener on the server spawns dedicated server processes for each connection request. The IPC option tells the listener to route requests to IPC addresses. Use the Advanced page to configure Net8's advanced features. TNS Time Out sets the SQLNET.EXPIRE_TIME parameter value (default is 0). With Net8, this also enables dead-connection detection. A small probe packet is sent out at the interval specified to make sure that a client/server connection is alive. Client Registration ID registers a unique client identified during a client request, which is then used for an audit trail. The other two parameters are UNIX Signal Handling (default is NO) and Disable Out-of-Band Break (default is OFF). For more details on these two parameters, refer to Oracle Net8 Administrator's Guide in the Oracle8 documentation set.

UNIX signal handling When an event occurs in UNIX, a signal flags a process that executes the relevant process code. Because UNIX doesn't allow events to call more than one signal, it's possible that a signal may not be made properly, and a defunct or dead process may not be cleaned up. The BEQUEATH_DETACH parameter in the SQLNET.ORA profile turns UNIX signal handling off or on. The default value, NO, leaves signal handling on. Selecting Naming from the Profile branch's drop-down list enables you to use various naming methods to configure your client/server connections. The page in Figure 24.4 has three tabbed pages: Methods, Oracle Names, and External. Figure 24.4 : Use the Promote and Demote buttons on the Methods page to set the order of the naming convention used. Disabling out-of-band breaks The DISABLE_OOB parameter in the SQLNET.ORA profile is used to disable out-of-band breaks. This parameter's default value, OFF, keeps the out-of-band breaks on. The Methods page includes two list boxes: Available Methods and Selected Methods. Oracle and Net8 uses the Naming methods in a top-down hierarchy. You can choose from several networking methodologies: q TNSNAMES (Local Naming). When a connection request to Net8 is made, the program first looks for the presence of a local TNSNAMES.ORA file with the connection description parameters in the file. If this file isn't present, Net8 tries to use the next naming method in the list. q HOSTNAME (Host Naming). Host Naming is a simple way of resolving connection descriptors. It's used mainly for clients and servers with the TCP/IP protocol. Host name resolution is done by using Domain Name Services (DNS), network information services, or a host table with the relevant information. q ONAMES (Oracle Names). This centralized naming system is used generally in larger installations where you have many servers each running several instances of Oracle. This system is relatively easier to administer after it's set up correctly. You can use Net8 Assistant to configure the Oracle Names Server and ensure that all parameters are set up correctly. q CDS, NDS, and NIS (External Naming). In this method, an adapter of the External Naming Protocol is installed on the client side. You then use Net8 Assistant to add the name of the protocol and the directory path information (NAMES.DIRECTORY_PATH) in the client profile. Cell Directory Services (CDS), Novell Directory Services (NDS), and Network Information Services (NIS) are the external naming services supported by Net8. You use the Oracle Names page when your network configuration is using Oracle Names. This page enables you to set the following values: q Database Domain Name with a default value of WORLD. q Maximum Wait Each Attempt (in the Resolution Persistence section) with a default value of 15 (seconds) on Windows NT. Default values depend on the operating system. This parameter specifies how long an Oracle Names client waits for a response from an Oracle Names server before trying to re-establish the connection. q Attempts Per Name Server (in the Resolution Persistence section) with a default value of 1. This specifies the number of attempts each Oracle Names client will try in the list of preferred names server before allowing the operation to fail. q Maximum Open Connections (in the Performance section) with valid values between 3 and 64. This parameter specifies the maximum number of open connections an Oracle Names client may have at one time to an Oracle Names Server.

http://www.informit.com/content/0789716534/element_024.shtml (9 of 13) [26.05.2000 17:15:36]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8
q

Initial Preallocated Requests (in the Performance section) with valid values between 3 and 256. Oracle Names lets you configure a number of messages in the message pool of the client. These messages can be used for future requests to names servers.

The External page has these parameters available: q Cell Directory Service (CDS/DCE) Cell Address. If you're using Cell Directory Service, add a valid DCE prefix to this field. q Netware Directory Service (NDS) Name Context. Enter a valid NDS name context in this field if you're using Novell's NetWare Directory Service (NDS). Net8 uses the NDS name context to look for a service name. q Network Information Services (NIS). Net8 needs to know the location of the file containing the meta map of NIS to map attributes. Configuring Service Names The Service Names branch of the network tree enables you to create aliases or service identifiers for connecting Net8 clients to Oracle8 servers. The dialog box and sequence is exactly the same as detailed in the section "Using Net8 Easy Config." Configuring Oracle Names Servers The last option under the network tree, Oracle Names Servers, is used to configure Oracle Names. The administrative functions available for Oracle Names configuration include discovering an Oracle Names Server, Creating a Names Server, Reloading All Names Servers, Navigating Oracle Names Server, and other options. Using Net8 Easy Config To launch Net8 Easy Config, choose Start, and then Programs; then select Oracle for Windows NT/9x and Oracle Net8 Easy Config. The first screen is the Oracle Service Name Wizard (see Figure 24.5). You can use this wizard to create a new service, modify an existing service, and test existing services. Figure 24.5 : Use the Oracle Service Names Wizard to configure network services. Type in new service names Select an existing service

In the following example, you use this wizard to add a service named FinProd. The alias (or service name) FinProd refers to an Oracle instance (SID) called finance running on server STARSHIP. Verifying changes made in TNSNAMES.ORA with a text editor You can view the TNSNAMES.ORA file in a text editor after changing or adding any information through Net8 Easy Config. You can see a new entry for the finance alias you create in the "Add a service with Net8 Easy Config" steps. Add a service with Net8 Easy Config 1. Launch Net8 Easy Config. 2. The Existing Services list box lists all the service descriptors observed earlier in the default configuration file. To add a new service, click Add New Service and type FinProd in the Net Service Name text box. Click Next to move to the next dialog box. 3. Select your network protocol. (Figure 24.6 shows TCP/IP selected.) Click Next. Figure 24.6 : Select a network protocol for configuring Net8 clients. 4. Enter information about the database server name (Host Name) or IP address (see Figure 24.7). If the listener is set to listen on a port other than 1521, you can make the change here. Click Next. Figure 24.7 : Select a host name and listener port. 5. Add the database instance name in the next dialog box-in this case, finance (see Figure 24.8). Figure 24.8 : Enter the database SID. 6. Click Finish to save the service name (see Figure 24.9). However, it's recommended that you first test the service to see if it works (see Figure 24.10). If the test is successful, you can save this information. Figure 24.9 : Test the service to see if a connection can be made. Figure 24.10: Verify to see if the created service works.

http://www.informit.com/content/0789716534/element_024.shtml (10 of 13) [26.05.2000 17:15:36]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

Connecting to Other Products with ODBC

Open Database Connectivity (ODBC) is a programming interface developed by Microsoft for providing Windows client-side products connectivity to ODBC-compliant databases (Oracle, of course, is ODBC-compliant). To enable connectivity to ODBC databases, you must install an ODBC driver, which runs on "top" of Net8 drivers. In other words, you can install the ODBC drivers if you have client-side connectivity working with Net8. Oracle, as well as third-party vendors, make ODBC drivers for Oracle databases. Examples of products connecting to Oracle with ODBC are PowerBuilder (Professional Edition), Microsoft Access, Microsoft Visual Basic, and Seagate Crystal Reports. For this example, configure an ODBC connection on a PC running Windows NT 4.0. The ODBC configuration utility is found in Control Panel. (When you install a standard Oracle client such as Net8 or SQL*Plus 8.0, you also can access another shortcut for the utility from the Oracle for Windows N NT folder.) You also have to install the Oracle ODBC driver on your PC. (You need administrator access if you're using Windows NT.) Some Oracle tools come with ODBC drivers You may want to verify whether you have an ODBC driver already installed on your PC. The ODBC administrator will list an ODBC driver, if present. The ODBC configuration utility has several tabbed pages (see Figure 24.11). Figure 24.11: You need to configure ODBC on a PC to work with Oracle Server. Data source name that only the user who configures it can see Can be operated and seen by all users of the PC; also visible to Windows NT services Enables you to connect to a data provider; users with the same drivers installed can share it Lists all ODBC drivers installed on the PC When tracing is enabled, logs all calls made to the ODBC drivers Lists the core components for the 32-bit ODBC administrator Add an ODBC system data source name 1. From the Windows Start menu, select Settings and then Control Panel. Launch the ODBC tool. 2. On the System DSN page, click the Add button. The Create New Data Source dialog box appears, listing all the ODBC drivers installed locally (see Figure 24.12). Figure 24.12: Create a new data source in the 32-bit ODBC Administrator. 3. Either double-click the Oracle ODBC driver or select it and click Finish. 4. Enter relevant information in the Oracle8 ODBC Driver Setup dialog box (see Figure 24.13). The Data Source Name, Description, and userid can be anything relevant to you. Enter FinProd in the Service Name text box. Click OK. Figure 24.13: Enter the data source name, description, and userid. With this, you're done configuring your system DSN. As you can see in Figure 24.14, the DSN FinanceConnect is listed as a system data source. Figure 24.14: Within each ODBC front-end product, all you need to do is simply use the name FinanceConnect hereafter as needed.

Connecting to Multithreaded Servers

When a client program such as SQL*Plus connects to an Oracle server with Net8, a process is spawned on the server. This process can be dedicated or shared (multithreaded). By default, the connection is made by using a dedicated server. For most online transaction processing (OLTP) applications, data is retrieved into client applications, operations are performed on the data, and changes are saved. Thus, data processing shifts between the client and the server. It's not always advantageous in terms of system resources to use dedicated server processes. You should use a dedicated connection, however, for database administration activities and batch loads. If the server operating

http://www.informit.com/content/0789716534/element_024.shtml (11 of 13) [26.05.2000 17:15:36]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

system can support it, you can use multithreaded or shared server processes (MTS). With the MTS option, a number of small processes share the workload of dedicated server processes. Two processes-shared servers and shared dispatchers-are created in addition to the standard Oracle database server processes.

Setting Net8 Entries and Initialization Parameters

To set up the MTS option, certain initialization parameters must be configured in the database instance's configuration file (INITSID.ORA) on the server. (INITSID.ORA's location varies from installation to installation.) You can use any text editor, such as Windows Notepad or UNIX vi, to configure the parameters: q MTS_LISTENER_ADDRESS is the address at which the listener listens by using a specific protocol. The syntax is as follows: MTS_LISTENER_ADDRESS = " (addr) " addr is the address from the LISTENER.ORA file. It can take the following form: MTS_LISTENER_ADDRESS = "(ADDRESS= (PROTOCOL=TCP) (HOST=STARSHIP) (PORT=1521) )" MTS_SERVICE specifies a service name for the dispatcher. The syntax is as follows: MTS_SERVICE = DB_NAME Oracle Corporation recommends using a unique identifier-the instance's SID if possible. If this value isn't set, it defaults to the DB_NAME value. MTS_DISPATCHERS controls the number of dispatcher processes created at startup. You can specify the number of dispatchers for each network protocol supported. The number of MTS dispatchers depends on the capacity of the database server, number of connections per process, and number of connections per network protocol. As a rule of thumb, the more dispatchers, the better the database performance will be. The Oracle8 Server Administrator's Guide gives a formula for calculating the initial number of dispatchers: number of dispatchers = CEIL (maximum number of concurrent sessions/connections per dispatcher) "Connections per dispatcher" depends on the operating system. MTS_MAX_DISPATCHERS limits the maximum number of dispatcher processes, including those belonging to all network protocols running on the server. There also may be a setting on the operating system for limiting the number of server processes. The formula for calculating this parameter is as follows: Max number of dispatchers = (maximum number of concurrent sessions/connections per dispatcher) MTS_SERVERS sets the initial number of shared server processes. It's recommended that you use a lower setting because the additional server processes require startup and are automatically deallocated as the load drops. MTS_MAX_SERVERS sets a limit on the maximum number of shared server processes. SHARED_POOL_SIZE has to be increased by about 1KB per user connecting to an Oracle server with the MTS option.

q q

Prerequisites for shared servers The database server's operating system must support the MTS option. Oracle on that platform must also support the multithreaded server option. Oracle on most UNIX and VMS platforms supports the MTS option. The Oracle-to-Windows NT port doesn't support the MTS option.

Connecting to Shared Servers

To connect from client software to a database server running with the MTS option enabled, use regular aliases or services generated with Net8 Easy Config, Oracle Net8 Assistant, or both. The database instance has to be stopped and started after the preceding parameters are introduced in the initialization.

Managing Dispatchers
Oracle8 provides two dynamic views-V$DISPATCHER and V$QUEUE-to monitor load on the dispatcher

http://www.informit.com/content/0789716534/element_024.shtml (12 of 13) [26.05.2000 17:15:36]

informit.com -- Your Brain is Hungry. InformIT - Configuring and Using Net8 Features From: Using Oracle8

processes. The number of dispatcher processes can be increased or decreased depending on the load and number of connections. Idle dispatchers are automatically terminated, until the number reaches MTS_DISPATCHERS, which acts as a lower limit for dispatcher processes. You can monitor this process from Server Manager (in motif mode) on UNIX servers. SEE ALSO For a listing of other dynamic views,

Managing Shared Servers

You also can monitor shared servers by using Server Manager in motif mode on UNIX-based systems. Idle shared server processes also are terminated automatically, to a lower limit of MTS_SERVERS. < Back Contents Next >

Save to MyInformIT

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/content/0789716534/element_024.shtml (13 of 13) [26.05.2000 17:15:36]

informit.com -- Your Brain is Hungry. InformIT - Featured Authors - Megh Thakkar

You are here : Home : Featured Author : Megh Thakkar Featured Book Linux System Administration Megh Thakkar
Search Tips

Exact Phrase All Words

Megh Thakkar is the Director of Database Technologies at Quest Software in Australia. Previously, he worked as a technical specialist at Oracle Corporation. He holds a masters degree in computer science and a bachelors degree in electronics engineering. Megh also holds several industry vendor certifications, including OCP, MCSE, Novell Certified ECNE, SCO UNIX ACE, and he is a Lotus Notes Certified Consultant. He is a frequent presenter at Oracle OpenWorld and various international Oracle User Groups. Megh is the author of E-commerce Applications Using Oracle8i and Java from Scratch and SAMS Teach Yourself Oracle8i on Windows NT in 24 Hours. He has also co-authored several books, such as Migrating to Oracle8i, Special Edition Using Oracle8/8i, Oracle8 Server Unleashed, C++ Unleashed, COBOL Unleashed, Oracle Certified DBA and, Using Oracle8. Megh is a renowned Oracle specialist who has performed Oracle development, consulting, support, and DBA functions worldwide over the past ten years. 1. e-Commerce Applications Using Oracle8i and Java From Scratch 2. Sams Teach Yourself Oracle8i on Windows NT in 24 Hours

Linux System Administration guides the reader in the many intricacies of maintaining a secure, stable system.

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/author/author_16287.shtml [26.05.2000 17:15:48]

informit.com -- Your Brain is Hungry. InformIT - Oracle8 Server Unleashed - Database - Oracle

You are here : Home : Oracle8 Server Unleashed Featured Author

Oracle8 Server Unleashed

Exact Phrase All Words Search Tips

Megh Thakkar Megh Thakkar is the Director of Database Technologies at Quest Software in Australia. Previously, he worked as a technical specialist at Oracle Corporation. Megh is the author of E-Commerce Applications Using Java and Oracle 8I From Scratch (Que, 2000).

Author: Joe Greene Web Price: $49.99 US Publisher: Sams ISBN: 0672312077 Publication Date: 6/17/98 Pages: 960 Table of Contents Save to MyInformIT This book will cover all features of Oracle8 such as: Oracle8 Architecture and the Network Computing Architecture; Migrating or building from scratch, and scaling to fit your business; Programming and Administration; Going beyond the server: system designing, development, and analysis using Developer/2000, Designer/2000 and Discover 3.0; Oracle Enterprise Manager; Using complex data types within a database i.e. business objects, multimedia, Java applets, etc.; Improved replication, performance and backup/recovery; OLTP; Deploying databases on the Web. Table of Contents PART I - ORACLE8 SERVER CHAPTER 1 -ORACLE8 CHAPTER 2 -ORACLE DATABASE ARCHITECTURE CHAPTER 3 -ORACLE8 NETWORK COMPUTING CHAPTER 4 -ORACLE8 SERVERS AND OPTIONS PART II - INSTALLATION, UPGRADE, AND MIGRATION CHAPTER 5 -IMPLEMENTATION PLANNING--PC TO ENTERPRISE SERVER CHAPTER 6 -INSTALLATION AND UPGRADES CHAPTER 7 -MIGRATING FROM NON-ORACLE DATABASES CHAPTER 8 -COEXISTING WITH NON-ORACLE DATABASES PART III - ORACLE8 COMPONENTS AND OBJECTS CHAPTER 9 -ORACLE PROCESSES CHAPTER 10 -ORACLE MEMORY AREAS CHAPTER 11 -ORACLE FILES CHAPTER 12 -ORACLE DATABASE OBJECTS CHAPTER 13 -ORACLE SYSTEM AND OBJECT PRIVILEGES CHAPTER 14 -ROLES AND GRANTS CHAPTER 15 -OBJECT-ORIENTED EXTENSIONS IN ORACLE8 CHAPTER 16 -REPLICATION CHAPTER 17 -QUERY AND TRANSACTION PROCESSING CHAPTER 18 -SUPPLIED PL/SQL PACKAGES PART IV - ORACLE8 ADMINISTRATION CHAPTER 19 -ORACLE8 ADMINISTRATIVE REQUIREMENTS CHAPTER 20 -ROUTINE MAINTENANCE AND SCHEDULING TASKS CHAPTER 21 -ADMINISTRATION USING ENTERPRISE MANAGER CHAPTER 22 -ORACLE8 TOOLS CHAPTER 23 -BACKUP AND RECOVERY CHAPTER 24 -ORACLE8 DATABASE TUNING CHAPTER 25 -ORACLE8 APPLICATION TUNING PART V - ORACLE8 AND THE WEB

Featured Book Oracle8 Server Unleashed This book covers hot topics such as deploying on the Web; data warehousing; business objects; and integrating multimedia, HTML, Java apps, and unstructured text in databases.

http://www.informit.com/product/0672312077/ (1 of 2) [26.05.2000 17:16:02]

informit.com -- Your Brain is Hungry. InformIT - Oracle8 Server Unleashed - Database - Oracle

CHAPTER 26 -ODBC/JDBC CHAPTER 27 -ORACLE WEB APPLICATION SERVER PART VI - IMPLEMENTATION OF ORACLE8 CHAPTER 28 -ORACLE8 CARTRIDGES CHAPTER 29 -USING ORACLE8 OBJECTS IN PL/SQL AND SQL CHAPTER 30 -DATA WAREHOUSES CHAPTER 31 -LARGE ONLINE TRANSACTION PROCESSING SYSTEMS CHAPTER 32 -MEDIA AND COMPLEX DATA SERVERS CHAPTER 33 -WORKING WITH VERY LARGE DATABASES

Contact Us | Copyright, Terms & Conditions | Privacy Policy | Advertising Copyright 2000 InformIT. All rights reserved.

http://www.informit.com/product/0672312077/ (2 of 2) [26.05.2000 17:16:02]