2.

Background on
Data Management
Aspects of Data Management and an
Overview of Solutions used in Engineering
Applications
 Overview
• Basic Terms
– What is data, information, data management, a data
model, a schema, metadata, etc.?
• Aspects of Data Management
– What is the most relevant functionality required for
engineering data?
• Data Management Approaches
– Files and File Systems
– Database Management Systems

Schallehn: Data Management for Engineering Applications

 Terms: Data and Information

Knowledge

Pragmatics
Information represent message or
Information observation transmitted as some form
Decoding
Encoding

of matter or energy.
Semantics
Data represent encoded information
Data with a fixed syntax (structure) and
meaning (semantics) usable by
Syntax computers.

Tokens of an Alphabet

Schallehn: Data Management for Engineering Applications

 Term: Data Management
• Term Data often used referring to static aspect of describing
and storing facts
• Term Information more often refers to dynamic aspects such
as transmission and communication
• Within the scope of this lecture focus on data and how it is
managed

Data Management refers to activities and concepts

to store, retrieve, manipulate, exchange, and
data according to a fixed syntax and semantics.

Schallehn: Data Management for Engineering Applications

 Term: Data Model
A data model is a model that describes in an abstract way how data
is represented in an information system or a database management
system.

• A data model defines syntax and semantics, i.e.

– How can data be structured (syntax)
– What does this structure mean (semantics)
• Very generic term for many applications
– Programming languages have their data models (e.g. C++ and Java have
object-oriented data models)
– Conceptual design methods (e.g. ER, UML) represent a data model
– File formats either apply a data model (e.g. XML) or implement their own
– Database management systems implement data(base) models

Schallehn: Data Management for Engineering Applications

 Term: Schema
A database schema is a map of concepts and their relationships for
a specific universe of discourse. It describes the structure of a
database.

• Specific for a given application

• During information design design two main types
– Conceptual schema represents information on an implementation-
independent level, often using graphical notations
– Logical schema represents information according to data model of
implementation language or system
• Often erroneously confused with Data Model

Schallehn: Data Management for Engineering Applications

 Data Model vs. Schema

Data Model Data Schema

Independent of application Application specific

Describes concepts of the schema Describes concepts of the application

“Language” used to describe data Description of data

Meta-metadata Metadata

Property of a programming language, Property of a program, conceptual diagram,

modeling language, database management database , etc.
system, etc.

Schallehn: Data Management for Engineering Applications

 Terms: Data and Metadata
• “Data about data”
• In Engineering
– Data describing products directly (geometry, product structure, simulation data, etc.)
– Metadata describing how data is used (users, processes, relations, documents, etc.)
• May be organized in several levels of abstraction (data, metadata, meta-
metadata, etc.)

Last Name
Columns of a table
First Name with a associated data
Landis Peter 28 type and constraints
Age
Spielberg John 39
Jackson Steven 26 Student Record of a table in
a relational database

Data Metadata (Schema) Meta-Metadata (Data Model)

Schallehn: Data Management for Engineering Applications

 Aspects of Data Management
• Focus on most important aspects for Engineering
– Persistence
– Data Integration
– Independence
– Security
– Data Quality
– Concurrency Control

Schallehn: Data Management for Engineering Applications

 Persistence
• Term Data Management primarily focuses on mid- and long-term
persistent storage of data beyond process runtime or system up-time
– In file systems or databases on secondary storage (hard disk)
– As backups on tertiary storage (archives on tapes, discs)
• Contrary to transient storage in main memory (RAM) or CPU registers or
cache
• Persistence mechanisms requires mapping main memory structures
(freely accessible address space) to secondary storage structures (e.g.
blocks, cylinders, etc. for hard disk)
• Archiving on tertiary media
– To prevent data loss due to common secondary storage failures
– To free secondary storage resources from data that is not frequently used
– For long-time documentation purposes required due to legal requirements

Schallehn: Data Management for Engineering Applications

 Data Integration
• Ideal state of information system: each real-world object is represented
in information system exactly once and accessible via a uniform
interface
• Goal: avoid redundancy
– Possible cause of inconsistencies
– Storage consumption (minor problem)
• Problems
– In most companies/organizations many information systems are used with
overlapping functionality and data
– Heterogeneity: data maybe represented in different ways (e.g. different
data model, different schema, different interfaces)
– Distribution: data is stored in possibly many departments, subsidiaries, etc.
– Hard to control within complex information system infrastructures
– Integration of data in files/file systems requires strong regulations

Schallehn: Data Management for Engineering Applications

 Data Integration /2
• Materialized integration: store Application Application Application
integrated data physically exactly 1 2 3
once and within one system
– Databases sometimes used to enforce
integration
Database
– E.g. EDM/PDM system, Warehouse Data
Server
• Virtual integration: provide single
point of access and reconciliation
mechanism for data physically stored
in several systems Database
– Federated databases and mediators

Typical ideal of a Database System as

an integration platform for several
applications

Schallehn: Data Management for Engineering Applications

 Independence
• Application should be How is data
External External External accessed
independent of how data is Schema 1 Schema 1 Schema 1 from different
stored and accessed applications?

• Decoupling allows flexible

Logical Independence
changes of storage
mechanism (e.g. conceptual What is the structure of the
Conceptual overall database for all
changes, optimizations, Schema applications?
replacement of entire
storage system) Physical Independence
• Achieved by using How is the data organized
standards, e.g. Internal Schema on disk and how can
accesses be optimized?
– XML for content of files
– Query language SQL fro
3-Level Schema Architecture of Database
databases
Systems providing independence from
storage details

Schallehn: Data Management for Engineering Applications

 Security
• Data Management must provide measures to protect data from
manipulation or disclosure of valuable data
• Access control mechanisms commonly used
– Authentication: mechanism to identify users of a systems
– Authorization: mechanism to grant or revoke specific access rights to users
– Backup (see above): mechanisms to prevent data loss
• Typical threats to data
– Data loss or integrity violations
– Intrusion: un-authorized access by misused identity or compromised security
measures
– Data Leakage: disclosure by authorized users to third partys

Schallehn: Data Management for Engineering Applications

 Data Quality
• “Fitness for use” of data in a given application
• Depends on requirements of this one application
• Quality may refer to many properties classified as so-
called “Dimensions” of Data Quality
• Data Quality can be
– Measured
– Improved
– Managed

Schallehn: Data Management for Engineering Applications

 Data Quality: Dimensions

[Lecture: Data Quality, Veit Köppen]

Schallehn: Data Management for Engineering Applications
 Concurrency Control
• Data management must support concurrent access by multiple users
• Synchronization required to solve possible problems with consistency
– Pessimistic concurrency control avoids problems by blocking operations
leading to violations
– Optimistic concurrency control checks for inconsistencies after operations
were performed and tries to resolve possible issues
• Synchronization mechanisms include
– Locking and exclusive accesses (pessimistic)
– Timestamp-based synchronization (pessimistic)
– Versions and variants (optimistic)
• Optimistic concurrency control often more suitable for engineering
application due to long-running interactive sessions

Schallehn: Data Management for Engineering Applications

 Data Management Approaches
• Data Management in Files
– File Systems
– File Formats
• Data Management in Databases
– Overview
– Relational DBMS
– Beyond Relational DBMS

Schallehn: Data Management for Engineering Applications

 File Systems
• Part of operating system to fulfill basic storage requirements
– Store units of data belonging together in an application context in files
– Semantics (format, interpretation) of file content is irrelevant to operating
system
– Content type often indicated by file type extension as part of the file name
– File system provides overlay structure of (hierarchical) directories/folders to
organize files and make them more easily accesible
– Operating systems maps these logical structures to physical of storage media
– Current operating systems support access control on in multi-user settings
• Distributed file systems used to connect physical resources (hard disks)
across nodes in a network and
– provide shared access typically within local network settings

Schallehn: Data Management for Engineering Applications

 Text vs. Binary Formats
ISO-10303-21;
• Text files HEADER;
/* Exchange file generated using ST-DEVELOPER 1.6 */
– Sequence of encoded
FILE_NAME(
characters, typically ASCII or /* name */ '1797609in',
/* time_stamp */ '1999-02-11T15:33:57-05:00',
UTF /* organization */ ('STEP Tools, Inc.'),
…
– Advantages
DATA;
• Content may be human-readable #10=ORIENTED_EDGE('',*,*,#11,.T.);
• Standardized encoding #11=EDGE_CURVE('',#13,#13,#12,.T.);
#12=INTERSECTION_CURVE('',#2476,(#2340,#2471),.CURVE_3D.);
…

• Binary Files
– Data encoded as variably used
sequence of bytes
– Advantages:
• Efficient storage usage
• Content protected/hidden to some
degree

Schallehn: Data Management for Engineering Applications

 Standard vs. Proprietary Formats
• Standardized file formats:
– Typically created or adopted by consortium or initiative including several
companies and/or organizations
– Examples in Engineering: STEP, IGES, JT, etc.
– Advantages:
• Allow easy data exchange between applications
• Useful for rachiving
• Proprietary formats:
– Formats developed for specific application by one developer
– Examples in Engineering: DWG and DXF (both AutoCAD), BRD (Eagle)
– “Industry standard”, if format used by several applications (e.g. for exchange)
and specification is publically available
– Advantages:
• More efficient
• Tailor-made for application (cover even specific functionality)

Schallehn: Data Management for Engineering Applications

 XML
• eXtensible Markup Language
– Hierarchical structure of nested elements (tags)
– Elements may have attributes
– Actual data on the leave level
– Mix of content (data) and description (schema, metadata)
• Developed based on SGML (document processing) to exchange any kind of
data on the Web
• Inspired by HTML (also based on SGML), which is only useful to exchange
• Can be considered a neutral text format for files
• Application-specific schemas of valid documents can be defined by
Document Type Definitions (DTD) or XML Shema (XSD)
• Standard software/libraries for XML processing publically available

Schallehn: Data Management for Engineering Applications

 XML Example: EAGLE .sch File
<schematic>
<parts>
<part name="SUPPLY1" deviceset="GND" device=""/>
<part name="C1" deviceset="C-EU" device="050-024X044" value="22pF"/>
</parts>
<sheets>
<sheet>
<instances> <!-- Positions the parts on the board. E. g.: -->
<instance part="SUPPLY1" gate="GND" x="132.08" y="187.96"/>
<instance part="C1" x="-50.8" y="200.66" rot="R270"/>
</instances>
<nets>
<net name="N$1" class="0">
<segment>
<wire x1="9.44" y1="19.04" x2="8.9" y2="19.04" width="0.15"/>
<wire x1="8.9" y1="19.04" x2="8.9" y2="20.66" width="0.15"/>
<wire x1="8.9" y1="20.66" x2="2.4" y2="20.66" width="0.15"/>
<pinref part="C1" pin="5"/>
<pinref part="SUPPLY1" pin="1"/>
</segment>
</net>
</nets>
</sheet>
</sheets>
</schematic> [Source: Philipp Ludwig]
 XML in Engineering
• Many formats based on XML
• Especially intended for data exchange
• Some examples:
– Collada for interactive 3D applications
– 3DXML for the exchange of geometrical data
– EAGLE board (BRD) and schema (SCH) files for electronic
circuits (see above)
– CAEX general purpose language for the exchange of
engineering data by European consortium
– AutomationML for plant engineering
– …
Schallehn: Data Management for Engineering Applications
 Database Systems
• Database Systems special solution for Applications Applications
high data management requirements
regarding
– Efficiency
– Consistency Database System
– Concurrent users
• Database (DB): collection of real world DBMS
facts within a given universe of discourse
persistently stored
• Database Management System (DBMS):
specific software independent of universe
of discourse
Database
• Database System (DBS): entire system of
database managed by DBMS

Schallehn: Data Management for Engineering Applications

 Database Management Systems
• Provide data to applications based on Client Server model
within a network
• Current systems mostly relational or object-relational (see
below)
– Oracle (object-relational, commercial)
– IBM DB2 (object-relational, commercial)
– Microsoft SQL Server (object-relational, commercial)
– PostgreSQL (object-relational, open source)
– MySQL (relational, open source)

Schallehn: Data Management for Engineering Applications

 Relational Database Systems
• Developed since early 1970s based on mathematical theory of
relations and operations performed on them (relational
algebra)
– Data is stored as records (tuples) in tables (relations) with values for
each column (attribute)
– Records can be identified by keys
– Keys can be used to establish connections across dat in different
tables (foreign keys)
– Constraints can be specified to grant consistency
• SQL (Structured Query Language) as a strong standard to
access relational databases

Schallehn: Data Management for Engineering Applications

 Relational Database Systems /2

PartID Name Weight SupplierID

GT-876-140425 Plunger 143.5 1
FT-852-130707 Shaft 77.0 3
FT-855-140809 Bolt 15.7 1
TT-707-778 Case 22.8 2

SupplierID Name Location

1 Reed & Sons New York
2 CaseStudio Boston
3 ToolTime Austin

Schallehn: Data Management for Engineering Applications

 Beyond RDBMS
• Object-Oriented DBMS (OODBMS or ODBMS): usage of
object-oriented concepts from OO-programming to add
semanticly rich modeling concepts
• Object-Relational DBMS (ORDBMS): integrate OO concepts
with relational model
• XML- and Document-Oriented DBMS: using semi-structured,
text-based formats for mass data storage
• NoSQL DBMS: flexible data models, mostly used within Cloud
context

Schallehn: Data Management for Engineering Applications