Tools and Techniques for Requirements Engineering and Management - Requirements Engineering and Management for Software Development Projects - page 173

subjected to integration testing. All defects uncovered during the tests are fixed.
The deliverable of this phase is the working source code and executable code.

13.2.6 Testing

In this phase, the executable code is subjected to system testing on the target
system on which the software is to be implemented and any defects uncovered are
fixed. The deliverable of this phase is the readiness of the system to roll into
piloting, parallel runs and production.

13.2.7 Delivery and Implementation

During this phase, the software is delivered to the target system and is imple-
mented, including conducting pilot runs of the system, running it in parallel with
the existing system and then finally rolling out the new system into production.
This phase also includes the preparation of user documentation including, the user
manual, operations manual and troubleshooting manual.

13.2.8 Software Maintenance

The next activity on software is obviously the maintenance of the software in
production. Maintenance includes bug fixing, modification and functional expan-
sion. The software would be replaced when a major change to the system becomes
necessary from the functionality point of view or due to a major technological
change such as obsolescence of the existing hardware or the onset of a phenom-
enon like the Y2K or the Internet.
SSADM also devised a few tools for modeling and documenting the require-
ments and design. These are:

1. Logical Data Modeling—It is the representation of a system which can be

either a manual system or a computerized system. During requirements
analysis, we build a logical data model of the manual system being taken up.
During design, we build a logical data model for the proposed computerized
system. Logical data model was proposed during the days of Hierarchical
DBMS (Database Management) and Network DBMS. With the onset of
Relational DBMS, ER (Entity Relationship) modeling was developed from the
logical data modeling.

2. Dataflow Modeling—It is building a model of the flow of data in the system.

As computerized systems basically process data, the system processes revolve