Transcript IMS1907 Database Systems

IMS1907 Database Systems
Summer Semester 2004/2005
Lecture 3
Database System Development and the
SDLC
Database Systems Development
Databases are key components of information systems
The development of the database must be coordinated with
all other activities in the development lifecycle
Database development requires specialised skills and
knowledge
Like IS development, database development requires a
structured approach
Monash University 2004
2
Database Systems Development
Database development requires a focus on the information
needs of a business
Information Engineering (IE) is a popular, data-oriented
methodology used to develop database systems
– data are modelled in the organisational context, not in the
usage, processing or technology context
– business context changes slowly  stable databases
– top-down planning
Monash University 2004
3
Database Systems Development
Top-down planning
– specific IS needs are deduced from understanding of
information needs
– broad perspective
– useful for considering integration of system components
– understanding of relationship between IS and business
objectives
– understanding of the impact of IS across organisation
Monash University 2004
4
Database Systems Planning
IE Planning phase
– goal is to align information technology and it’s usage with
the overall strategic goals of the organisation
– alignment is essential to achieving maximum benefits
from the investment in technology
– aims at an ‘enterprise’ view of the information needs of
an organisation
– three steps in the phase
Monash University 2004
5
Database Systems Planning
The three steps in the IE Planning phase
– identify strategic planning factors
– identify corporate planning objects
– develop an enterprise model
Monash University 2004
6
Database Systems Planning
Step 1 - identify strategic planning factors
– goals
– critical success factors (CSF)
– problem areas
• see Hoffer, Prescott and McFadden, (2005), Table 2-2, p. 41
Identifying these factors enables
– the development of planning context
– the linkage of IS plans with strategic business plans
– setting of priorities for new IS requests
Monash University 2004
7
Database Systems Planning
Step 2 - identify corporate planning objects
– organisational units
– organisational locations
– business functions
– entity types
– information systems
• see Hoffer et al, (2005), Table 2-3, p. 42
Defines business scope and where IS changes can occur
Monash University 2004
8
Database Systems Planning
Step 3 – develop an Enterprise Model
– functional decomposition of each business function
– enterprise data model
– various planning matrices
• see Hoffer et al, (2005), Figure 2-3, p. 44
Helps simplify problems, isolate attention
Identify business rules
Setting development priorities, scheduling activities
Monash University 2004
9
Database Systems Planning
Planning matrices
– location-to-function
– unit-to-function
– information system-to-data entity
– supporting function-to-data entity
– information system-to-objective
Identifying orphans, missing entities, missing functions,
unassigned functions, unassigned units, necessary
systems, prioritisation of development
Monash University 2004
10
Database Systems Development
Not all database systems arise from a top-down planning
approach
Bottom-up requests can cause a need for development
– operational level requests
– projects requested by IS users to perform job
– need for data management improvements
There is still a need for an enterprise model of data
– data already exists? new data requirements? more than
one database?
Monash University 2004
11
Database Development and the SDLC
Initiation
Enterprise modelling
Conceptual data modelling
Analysis
Logical database design
Design
Physical database design
and definition
Implementation
Database implementation
Review
Maintenance
Database review
Database maintenance
Monash University 2004
12
Enterprise Modelling
Review enterprise modelling components identified during
planning
Analyse current IS, database and data processing
Analyse general business functions and data needs
Describe new information and data needs
Determine which data already exists
Justify need for new data and databases to support
business
Monash University 2004
13
Enterprise Data Model
High-level view of major ‘things’ of significance to the
organisation
Similar to entity-relationship modelling but not as detailed
Business-oriented descriptions of elements
Statements of business rules governing data validity
Monash University 2004
14
Enterprise Data Model
A possible simplified Enterprise Data Model for
Amazon.com
BOOK
CUSTOMER
SALE
CD
Monash University 2004
15
Conceptual Data Modelling
Identify scope of database requirements
Analyse overall data requirements to support functionality
Develop preliminary data model - entity-relationship (ER)
modelling
Compare conceptual ER model with enterprise data model
Develop detailed conceptual data model – entities,
relationships, attributes, and business rules
Make conceptual model consistent with other IS models
Populate repository with all conceptual DB specifications
Monash University 2004
16
Logical Database Design
Transform conceptual model into logical data model
– analyse in detail transactions, forms, displays and
enquiries (DB view) needed to support functions
– integrate database views and newly discovered
requirements into conceptual model
– identify data integrity and security requirements
– transform reconciled data specifications into stable data
structures – dependent on type of DBMS
Start to specify logic for maintaining and querying database
Populate repository
Monash University 2004
17
Physical Database Design and Definition
Requires knowledge of specific DBMS used
Define database to DBMS (often generated by repository)
Decide on physical organisation of data – records, file
organisation, indexes, clustering
Design database processing programs necessary to
generate information
Enables secure and efficient handling of data processing
needs
Coordinated with design of other IS components – programs,
hardware, operating systems, networks
Monash University 2004
18
Database Implementation
Code, test and install database processing programs
Complete database documentation and training materials
Put procedures in place for ongoing support of DB and IS
Install database
Load and convert data from legacy systems
Load any new data needed
Put database into production
Monash University 2004
19
Database Maintenance
Analyse database and database applications to ensure
evolving information needs are met
Tune database for optimum performance
Fix errors in database and database applications
Recover or rebuild database if corrupted or contaminated
due to program or system malfunction or failure
Typically the longest step in DB development – lasts
throughout the life of the database and associated
applications
Monash University 2004
20
Packaged Data Models
Reuse of standard, but flexible, proven data models
Can save time in modelling data requirements
Comparatively low cost
Can be customised and incorporated into other data models
Developed by industry specialists and DBMS vendors
Based on experience and expertise across industry sectors
Two principal types of packaged data models
– universal data models
– industry-specific data models
Monash University 2004
21
Packaged Data Models
Universal data models
– core subject areas common to many businesses –
customers, products, accounts, documents, projects
– core functions common to businesses that follow similar
patterns – purchasing, accounting, receiving, PM
Provide templates for one or more of these areas
Based on the fact that although differing in detail, underlying
data structures are similar
Monash University 2004
22
Packaged Data Models
Industry-specific data models
– generic data models for use in specific industry area
– available for nearly every major industry group – health
care, telecommunications, discrete manufacturing,
process manufacturing, banking, insurance, mining, etc
• see Hoffer et al, (2005), Figure 2-7, p. 51
Based on fact that process and data needs are similar within
industry, but can differ across industries
Monash University 2004
23
People Involved in Database Development
Systems Analysts
– analyse business situation
– identify business needs to meet problems or
opportunities
Database Analysts
– determine requirements for database
– design database
Users
– provide assessment of information needs
– monitor that system meets their requirements and needs
Monash University 2004
24
People Involved in Database Development
Programmers
– design and write programs to maintain and access data
Data and Database Administrators
– responsibility for existing and future databases
– ensure consistency and integrity across databases
– expert consulting and training
Other technical experts
– networks, operating systems, communications, testing,
documentation
Monash University 2004
25
References
Elmasri, R. and Navathe, S.B., (2000), Fundamentals of
Database Systems, (3rd edn.), Addison-Wesley, Reading,
Massachusetts, USA.
Hoffer, J.A., Prescott, M.B. and McFadden, F.R., (2005),
Modern Database Management, (7th edn.), Pearson
Education Inc., Upper Saddle River, NJ, USA.
Monash University 2004
26