UNIT-01

Introduction

 Overview:

o A Database Management System (DBMS) is a software system designed to

manage and organize data in a structured manner, making it efficient to store,
retrieve, update, and delete data.

o It provides a controlled environment for data access and manipulation,

ensuring data integrity, security, and concurrency.

 Database System vs. File System:

o File System:

 Collection of files stored on a storage device.

 Data is stored in individual files with limited relationships between

them.

 Limited data integrity and security features.

o Database System:

 Centralized repository of data with well-defined relationships between

different data items.

 Provides mechanisms for data integrity, security, concurrency control,

and recovery.

 Offers a higher level of abstraction and data independence.

 Database System Concept and Architecture:

o Three-Schema Architecture:

 External Schema: Individual user views of the database.

 Conceptual Schema: An abstract representation of the entire

database.

 Internal Schema: How data is physically stored on the storage devices.

 Data Model, Schema, and Instances:

o Data Model: A collection of concepts that describe the structure of a database.

(e.g., Relational Model, Entity-Relationship Model)
 o Schema: A specific description of a particular database within a given data
model.

o Instance: A specific set of data stored in the database at a particular point in

time.

 Data Independence and Database Language and Interfaces:

o Data Independence: The ability to change the schema at one level of the
database system without affecting the schema at the next higher level. 1

 Logical Data Independence: Changes in the conceptual schema do not

affect external schemas.

 Physical Data Independence: Changes in the internal schema do not

affect the conceptual schema.

o Database Languages:

 Data Definition Language (DDL): Used to define the database schema

(e.g., CREATE, ALTER, DROP).

 Data Manipulation Language (DML): Used to manipulate data within

the database (e.g., SELECT, INSERT, UPDATE, DELETE).

 Overall Database Structure:

o Typically consists of tables (relations) with rows (tuples) and columns

(attributes).

o Relationships between tables are defined using primary keys and foreign keys.

Data Modeling Using the Entity-Relationship Model (ER Model)

 ER Model Concepts:

o Entity: A real-world object or concept that can be uniquely identified (e.g.,

Student, Department, Course).

o Attribute: A property or characteristic of an entity (e.g., StudentID, Name,

Department).

o Relationship: An association between two or more entities (e.g., Student

'enrolls in' Course).

 Notation for ER Diagram:

o Entity: Represented by a rectangle.

 o Attribute: Represented by an oval or circle.

o Relationship: Represented by a diamond.

o Cardinality: Represents the number of instances of one entity that can be

associated with instances of another entity (e.g., 1:1, 1:N, N:M).

 Mapping Constraints:

o One-to-One (1:1): One instance of entity A is associated with at most one

instance of entity B, and vice versa.

o One-to-Many (1:N): One instance of entity A is associated with many instances

of entity B, but one instance of entity B is associated with at most one instance
of entity A.

o Many-to-Many (N:M): One instance of entity A is associated with many

instances of entity B, and one instance of entity B is associated with many
instances of entity A.

 Keys:

o Super Key: A set of attributes that uniquely identifies each tuple in a relation.

o Candidate Key: A minimal super key (no subset of the attributes can uniquely
identify tuples).

o Primary Key: A candidate key chosen to uniquely identify each tuple in a

relation.

 Generalization:

o Represents a 'is-a' relationship between entities.

o A more general entity is called a superclass, and more specific entities are
called subclasses.

 Aggregation:

o Represents a 'part-of' or 'has-a' relationship.

o Treats a group of entities as a single, higher-level entity.

 Reduction of an ER Diagram to Tables:

o Translate entities and relationships into tables.

o Represent attributes as columns in tables.

o Use primary and foreign keys to represent relationships between tables.

 Extended ER Model:
 o Includes additional features like weak entities, subtypes, and
specialization/generalization hierarchies.

 Relationship of Higher Degree:

o Represents a relationship between more than two entities.

o Can be modeled using additional tables or by decomposing the relationship

into binary relationships.

This provides a comprehensive introduction to database systems, including the ER

model and its key concepts.

UNIT-02

Relational Data Model

 Concepts:

o Relation: A named, two-dimensional table with rows (tuples) and columns

(attributes).

o Tuple: A single row in a relation, representing a record.

o Attribute: A named column in a relation, representing a specific property of an

entity.

o Domain: A set of permitted values for an attribute.

 Integrity Constraints:

o Entity Integrity:

 Primary Key: Uniquely identifies each row in a table.

 No attribute of a primary key can be NULL.

o Referential Integrity:

 Foreign Key: A column in one table that references the primary key of
another table.

 Ensures that related data exists and maintains consistency between

tables.

o Domain Constraints: Define the valid values for each attribute.

 o Key Constraints: Primary Key, Unique Key, Candidate Key, Super Key.

 Relational Algebra:

o A set of operations to manipulate relations:

 Selection (σ): Selects rows that satisfy a given condition.

 Projection (π): Selects specific columns.

 Union (∪): Combines rows from two relations.

 Intersection (∩): Selects rows that exist in both relations.

 Difference (-): Selects rows that exist in one relation but not the other.

 Cartesian Product (×): Combines all rows from one relation with all
rows from another.

 Join (⨝): Combines rows from two relations based on a matching

condition.

 Relational Calculus:

o Tuple Relational Calculus (TRC): Defines a query as a set of tuples that satisfy
a given formula.

o Domain Relational Calculus (DRC): Defines a query as a set of domain values

that satisfy a given formula.

Introduction to SQL

 Structured Query Language: A powerful language for managing relational databases.

 Characteristics:

o Declarative language: Specifies what data to retrieve, not how to retrieve it.

o High-level language: Easy to learn and use.

o Standard language: Supported by most relational database systems.

 Advantages:

o Data independence: Changes in data storage do not affect SQL queries.

o High-level abstraction: Simplifies complex data manipulation tasks.

o Powerful and flexible: Can perform a wide range of operations.

SQL Data Types and Literals

 Data Types:
 o Numeric (INTEGER, FLOAT, DECIMAL)

o Character (CHAR, VARCHAR, TEXT)

o Date/Time (DATE, TIME, TIMESTAMP)

o Boolean (BOOLEAN)

o Binary (BLOB)

 Literals:

o Constant values used in SQL statements.

o Example: 'John Doe', 123, '2023-11-28'

Types of SQL Commands

 DDL (Data Definition Language):

o CREATE: Creates database objects (tables, views, indexes).

o ALTER: Modifies the structure of database objects.

o DROP: Deletes database objects.

 DML (Data Manipulation Language):

o SELECT: Retrieves data from tables.

o INSERT: Adds new rows to a table.

o UPDATE: Modifies existing rows in a table.

o DELETE: Removes rows from a table.

 DCL (Data Control Language):

o GRANT: Grants privileges to users.

o REVOKE: Revokes privileges from users.

SQL Operators and Their Procedure

 Arithmetic Operators: +, -, *, /

 Comparison Operators: =, <>, <, >, <=, >=

 Logical Operators: AND, OR, NOT

 Set Operators: UNION, INTERSECT, EXCEPT (MINUS)

Tables, Views, and Indexes

 Tables: The fundamental data storage unit in a relational database.

  Views: Virtual tables that provide a customized view of data from one or more tables.

 Indexes: Data structures that improve the speed of data retrieval.

Queries and Subqueries

 Queries: SELECT statements that retrieve data from the database.

 Subqueries: Nested SELECT statements within another SQL statement.

Aggregate Functions

 Functions that summarize data:

o AVG(): Calculates the average value.

o SUM(): Calculates the sum of values.

o COUNT(): Counts the number of rows.

o MIN(): Finds the minimum value.

o MAX(): Finds the maximum value.

Insert, Update, and Delete Operations

 INSERT: Adds new rows to a table.

 UPDATE: Modifies existing rows in a table.

 DELETE: Removes rows from a table.

Joins

 Combine data from multiple tables:

o INNER JOIN: Returns rows where there is a match in both tables.

o LEFT JOIN: Returns all rows from the left table and matching rows from the
right table.

o RIGHT JOIN: Returns all rows from the right table and matching rows from the
left table.

o FULL JOIN: Returns all rows from both tables.

Unions, Intersections, Minus

 Set operations that combine or compare results from multiple queries.

Cursors

 Allow row-by-row processing of a result set.

Triggers
  Automatically execute a block of code in response to a database event (e.g., INSERT,
UPDATE, DELETE).

Procedures in SQL/PL SQL

 Stored blocks of code that can be executed multiple times.

 Improve code reusability and maintainability.

This comprehensive overview covers the essential concepts of the relational data model, SQL,
and related topics.

UNIT-03

Database Design & Normalization

1. Functional Dependencies (FDs)

 Definition: A functional dependency X → Y holds between two sets of attributes X and

Y in a relation R if, for every pair of tuples t1 and t2 in R, if t1[X] = t2[X], then t1[Y] =
t2[Y].

o In simpler terms: If you know the value of X, you can uniquely determine the
value of Y.

 Example: In a relation Employee(EmpID, Name, Dept), EmpID → Name holds because

each employee has a unique name associated with their ID.

2. Normal Forms

 1NF (First Normal Form):

o Atomic Values: Each attribute contains only a single, atomic value (no lists or
repeating groups within a single cell).

o No Repeating Groups: Each row must have the same number of columns.

 2NF (Second Normal Form):

o In 1NF.

o Full Functional Dependency: Every non-key attribute is fully functionally

dependent on the entire primary key.
  No partial dependencies: Non-key attributes should not depend on only
a part of the primary key if it's composite.

 3NF (Third Normal Form):

o In 2NF.

o No Transitive Dependencies: Non-key attributes should not depend on other

non-key attributes.

 BCNF (Boyce-Codd Normal Form):

o A stricter version of 3NF.

o Every determinant is a candidate key: If X → Y holds, then X must be a

superkey (a set of attributes that uniquely identify a tuple).

 Higher Normal Forms (4NF, 5NF):

o Address more complex dependencies like multi-valued dependencies (4NF)

and join dependencies (5NF).

3. Inclusion Dependencies

 Constraints that specify that a subset of tuples in one relation must also appear in
another relation.

4. Lossless Join Decompositions

 A decomposition of a relation into multiple relations is lossless if the original relation

can be perfectly reconstructed by joining the decomposed relations.

5. Normalization Using FDs, MVDs, and JDs

 FDs (Functional Dependencies): Used to achieve 1NF, 2NF, 3NF, and BCNF.

 MVDs (Multi-valued Dependencies): Used to achieve 4NF.

o Express that multiple values of one attribute are associated with a single value
of another attribute, independent of other attributes.

 JDs (Join Dependencies): Used to achieve 5NF.

o Express that a relation can be reconstructed by joining a set of its projections.

6. Alternative Approaches to Database Design

 Denormalization: Intentionally introducing redundancy to improve performance in

specific cases (e.g., for frequently accessed data).

 Object-Oriented Databases: Store data as objects with attributes and methods, better
suited for complex data structures.
  NoSQL Databases: Designed to handle large volumes of unstructured or semi-
structured data (e.g., JSON, XML).

Key Goals of Normalization

 Minimize Redundancy: Reduce data duplication to improve data integrity and

efficiency.

 Reduce Anomalies: Prevent insertion, deletion, and update anomalies that can occur
due to redundancy.

 Improve Data Integrity: Ensure the accuracy and consistency of data.

 Improve Data Maintainability: Make it easier to modify and update the database
schema.

Normalization Process

1. Identify FDs, MVDs, and JDs: Analyze the data requirements and identify the
dependencies between attributes.

2. Decompose Relations: Break down relations into smaller relations to achieve the
desired normal form.

3. Evaluate Decompositions: Ensure that the decompositions are lossless and

dependency-preserving.

By following these principles, database designers can create well-structured and efficient
databases that meet the specific needs of an application.

UNIT-04

Transaction Processing Concept

 Transaction: A logical unit of work that accesses and possibly modifies
data in a database.
 ACID Properties:
o Atomicity: All-or-nothing execution. A transaction either completes
entirely or has no effect.
o Consistency: Preserves database integrity constraints.
 o Isolation: Concurrent transactions appear to execute serially.
o Durability: Once committed, changes are permanent and survive
system failures.
Testing of Serializability
 Serializability: A schedule is serializable if its effect on the database is
equivalent to the effect of executing the transactions in some serial order.
 Methods:
o Precedence Graph: A directed graph to visualize the order of
conflicting operations. If the graph has no cycles, the schedule is
conflict serializable.
o View Serializability: A schedule is view serializable if it is equivalent
to some serial schedule with respect to the final state of the
database and the values read by each transaction.
Conflict & View Serializable Schedule
 Conflict Serializable: A schedule is conflict serializable if it is equivalent to
some serial schedule with respect to the order of conflicting operations.
 View Serializable: A schedule is view serializable if it is equivalent to some
serial schedule with respect to the final state of the database and the
values read by each transaction.
 Relationship: Conflict serializability is a stronger condition than view
serializability. All conflict serializable schedules are view serializable, but
not all view serializable schedules are conflict serializable.
Recoverability
 Recoverability: Ensures that if a transaction fails, the effects of its partial
execution are undone, and the database is restored to a consistent state.
 Types:
o Cascadeless Recovery: A transaction can only commit if all
transactions it read from have already committed.
 o Strict Recovery: A transaction can only commit if all transactions it
read from and wrote to have already committed.
Recovery from Transaction Failures
 Log-Based Recovery:
o Log: A sequence of records that record the actions of transactions.
o Types of Log Records:
 Start: Indicates the beginning of a transaction.
 Commit: Indicates the successful completion of a
transaction.
 Abort: Indicates the termination of a transaction due to
failure.
 Write: Records the value written to a data item.
o Recovery Techniques:
 Undo: Reverts the effects of aborted transactions.
 Redo: Re-executes the effects of committed transactions
that were not written to the stable storage before a system
crash.
Checkpoints
 Checkpoints: Periodically recorded points in the execution of
transactions.
 Purpose:
o Reduce recovery time after a system crash.
o Minimize the amount of log that needs to be processed during
recovery.
Deadlock Handling
 Deadlock: A situation where two or more transactions are waiting for
resources held by each other, resulting in a stalemate.
  Detection:
o Wait-for Graph: A directed graph to represent the wait-for
relationships between transactions. If the graph contains a cycle, a
deadlock exists.
 Prevention:
o Timestamp Ordering: Assign timestamps to transactions and
enforce an order based on timestamps.
o Resource Ordering: Impose a partial order on resources and
require transactions to request resources in that order.
 Resolution:
o Timeout: Terminate a transaction that has been waiting for
resources for too long.
o Deadlock Detection and Rollback: Detect deadlocks and abort one
or more transactions involved in the deadlock.
Distributed Database
 Distributed Data Storage: Data is physically stored on multiple computers
in a network.
 Advantages:
o High Availability: Increased fault tolerance.
o Scalability: Easy to add more nodes to the system.
o Local Autonomy: Each site can manage its own data.
 Concurrency Control: Techniques to ensure data consistency in a
distributed environment.
o Two-Phase Locking (2PL): Each transaction acquires all locks before
releasing any locks.
o Timestamp Ordering: Transactions are assigned timestamps, and
operations are executed in timestamp order.
 o Optimistic Concurrency Control: Transactions execute without any
restrictions, and conflicts are detected and resolved at commit
time.
Directory System
 Maintains information about the location of data fragments in the
distributed database.
 Enables efficient data access and query processing.
 Types:
o Centralized Directory: A single node maintains the directory
information.
o Distributed Directory: Directory information is partitioned and
replicated across multiple nodes.
This comprehensive overview covers the key concepts of transaction processing,
concurrency control, and distributed databases.

UNIT-05

Concurrency Control
 Definition: Mechanisms to ensure that concurrent execution of
transactions in a database system preserves data consistency and avoids
anomalies like lost updates, dirty reads, and uncommitted dependency.
Locking Techniques
 Two-Phase Locking (2PL):
o Growing Phase: Transactions acquire locks before accessing data
items.
o Shrinking Phase: Transactions release locks without acquiring any
new locks.
 o Strict 2PL: A transaction cannot release any lock until it commits or
aborts.
 Lock Compatibility Matrix: Defines which locks can be granted
concurrently on the same data item.
 Granularity of Locking:
o Fine-grained Locking: Locks are acquired on individual data items
(e.g., tuples, records).
o Coarse-grained Locking: Locks are acquired on larger units (e.g.,
tables, pages).
Timestamping Protocols
 Basic Timestamp Ordering: Each transaction is assigned a unique
timestamp. Transactions are executed in timestamp order.
 Thomas Write Rule: Allows a later timestamp transaction to overwrite an
earlier timestamp transaction's write if the earlier transaction has already
committed.
Validation Based Protocol
 Optimistic Concurrency Control: Assumes that conflicts are rare.
 Phases:
o Read Phase: Transactions read data items without acquiring locks.
o Validation Phase: Before committing, the transaction checks for
conflicts with other committed transactions.
o Write Phase: If no conflicts are detected, the transaction writes its
changes to the database.
Multiple Granularity
 Hierarchical Locking: Data is organized in a hierarchy (e.g., database,
table, page, tuple).
 Intention Locks: Indicate the intention to acquire locks at finer granularity
levels.
  Example: A transaction may acquire an intention lock on a table before
acquiring exclusive locks on specific tuples within that table.
Multi-Version Schemes
 Maintain multiple versions of data items.
 Allow concurrent readers to access older versions of data while writers
create new versions.
 Reduces blocking and improves concurrency.
Recovery with Concurrent Transactions
 Log-Based Recovery:
o Undo/Redo Logs: Record the actions of transactions.
o Checkpointing: Periodically record the state of the system to
reduce recovery time.
 Concurrency Control and Recovery:
o Concurrency control mechanisms must be coordinated with
recovery techniques to ensure consistent database state after
failures.
Case Study of Oracle
 Oracle Database: A popular commercial database system that employs a
combination of concurrency control techniques, including:
o Multi-Version Concurrency Control (MVCC)
o Locking (row-level, table-level)
o Granularity Control

 Oracle also incorporates advanced features like:

o Parallel Execution
o Data Warehousing
o High Availability and Disaster Recovery
This provides a comprehensive overview of concurrency control techniques in
database systems.