Transcript ppt - DUET

Introduction to
Database Management
Systems
Lila Rao Graham
1
Motivation
 Organisations are increasingly aware of the
importance of information in the solution of
their problems.
 Because of decreasing cost of data storage,
organisations store increasing quantities of
data.
 This data must be managed in the most
efficient and effective manner.
2
The Database Approach
 Organizations must have access to
operational data that is

accurate




timely
convenient
up-to-date
secure but available
 As control decentralizes in an organization,
there is a danger that data management
decentralizes as well.
3
Dangers of Decentralized Data
Management
 Redundancy
 Inconsistency
 Incompatibility



Different formats
Different constraints
Different models of the data
 Inaccuracy
 Insecurity
 Non-accountability
 Inflexibility
4
Integration and Sharing
 Data should be

integrated


the database can be regarded as a unification of
several otherwise distinct data files
shared

individual pieces of data in the database can be
shared among several different users
5
Data Management
 Objective

Model a part of the world of the real world
accurately and efficiently to give user useful
data
 Requirements





Flexibility
Accuracy
Reliability
Accessibility
Efficiency
6
Data Modelling: Abstraction Again
 Abstraction:

In modelling the data, we again use
abstraction. We hide (irrelevant) detail and
concentrate on the general and common
attributes of a set of objects
 Example:

Real world:


Student
Model:

ST(id, name, dob, grade history)
7
Definition of a Database
 A database (DB) is a collection of interrelated
computer files, whose data contents and
structure are described in a data dictionary
and which is under the control of a database
management system (DBMS)
8
Database Management Systems
 A database management system is a
collection of programs that carry out activities
on a database, including




setting up storage structures
loading data
accepting and performing updates
accepting data requests from users and
programs.
9
Functions of a DBMS
 A good DBMS performs the following
functions







maintain data dictionary
support multiple views of data
enforce integrity constraints
enforce access constraints
support concurrency control
support backup and recovery procedures
support logical transactions
10
Advantages of Using a DBMS
 All applications access a common DB. So,
details related to data storage and access
are removed from programs and users.
 Hence






less redundancy
less risk of inconsistency
maintenance of data integrity
application of access restrictions
balance between different requirements
data independence
11
Layered Structure of a Database
 We can regard a database as having a
layered structure
-----------------------------External Schema
----------------------------------------------------------Conceptual Schema
----------------------------------------------------------Internal Schema
------------------------------
12
Schemas and Instances
 A schema is the overall design of the
database. It describes the data contents,
structure and some other aspects of the
database also called the intension of the
database
 The instance is the collection of data stored in
the database at a particular time, also called
the extension of the database.
13
Data Independence:
External schema
 Different users often need different views of
the data.
 Example:


Bursar needs to have access to financial
information on a student
Head of Department needs access to
academic information.
 An external schema is a description of part of
the DB as seen by an application
programmer or a user
14
Conceptual Schema
 Representation of the logical structure of the
information content of the DB.
 Abstracts away from the actual physical
storage.
 The data “as it really is”.
 It is, in a sense, a composite of all the
external schemata.
15
Internal Schema
 The internal schema describes the data as it
is physically stored.
 For example,



record structure
types of fields in a record
existence of primary and secondary indexes
16
Data Independence
 Higher levels should be immune to changes
at the lower level.
 Physical data independence:

changes at the internal schema (e.g., changes
in record structure or indexes) do not affect
the conceptual schema
 Logical data independence

changes at the conceptual schema (e.g.,
addition of further fields in a record) do not
affect the external schema.
17
Importance of Data Independence
 Similarity between data independence and
the use of abstract data types.


Immunity to lower-level implementation
details.
Possibility of adding new applications without
having to restructure the data base.
18