IS 341

Business Systems Analysis

Chapter 10 - Notes

Systems Implementation and Operation

 

I.  Systems Implementation and Operation – the most expensive and time-consuming phase of the SDLC – expensive because so many people are involved, time-consuming because of the work that must be completed throughout the lifetime of the system;  the seven major activities of this phase fall into three groups:

                   1.  Activities that lead to the system going into operation

                             1.  Coding – turning physical design specifications into working computer code

                             2.  Testing – must be planned early in the process to decide what needs to be tested and to gather test data

                             3.  Installation – actually replacing the old system with the new one; 4 methodologies:

                   2.  Activities necessary for successful system operation

                             4.  Documentation – for both developers and users

                             5.  Training – users need training in the use and operation of a system;

                             6.  Support – where to go for help:  online (automated) or to a real person (help desk)

                   3.  Ongoing activities to keep the system working and up-to-date

                             7.  Maintenance – from the moment it is implemented, a system will need maintenance for errors, modifications (changing business needs and changing business environment), and requests for enhancements (new capabilities)

          A.  The Process of Coding, Testing, and Installation – (see above)

          B.  Deliverables and Outcomes from Coding, Testing, and Installation

                   1. Coding – Code and program documentation

                   2.  Testing – Test plan and test data, Results of testing

                   3.  Installation – User guides, User training plan, Installation and conversion plan

          C.  The Process of Documenting the System, Training Users, and Supporting Users – two groups of people will use the documentation:  systems personnel who maintain the system and users who work with the system; documentation is important to both groups

          D.  Deliverables and Outcomes from Documenting the System, Training Users, and Supporting Users

                   1.  Documentation – System documentation and user documentation

                   2.  Training – Classes and tutorials

          E.  The Process of Maintaining Information Systems – four major activities:

                   1.  Obtaining maintenance requests – should be a formal process , done in writing with signatures on documents

                   2.  Transforming (translating) request into changes – understand (sometimes decipher) what the user actually wants (not necessarily what he actually asked for)

                   3.  Designing changes – a mirror (sort of…) of the SDLC, planning just what is to be done

                   4.  Implementing changes – put the changes into effect, training, and documentation

          F.  Deliverables and Outcomes from Maintaining Information Systems – development of new versions of specific parts of the system, along with new documentation, training, etc.

 

II.  Software Application Testing – testing starts early in the SDLC process and continues throughout the SDLC; some firms have special groups whose responsibility is Testing

          A.  Seven Different Types of Testing (see Table 10-3, p. 369)

                   1.  Manual

                             a.  Inspections – physically looking at the code for errors, usually subtle errors not picked up by automated checks

                             b.  Walkthroughs – formal review of the code by one or more persons along with the coder

                             c.  Desk Checking – executing the program code one step at a time (manually ) by a reviewer

                   2.  Automated

                             a.  tests for specific errors in code grammar

                             b.  Unit (Module) Testing – test the modules individually for errors

                             c.  Integration Testing – bring the modules together for testing as a whole

                             d.  System Testing – bring together the entre set of programs to see if they work together as a whole

          B.  The Testing Process –testing is not haphazard or a last-minute thought before cutover; it must be planned and executed carefully to be effective

          C.  Acceptance Testing by Users – users test a completed system and give their acceptance of it

                   1.  Alpha Testing – testing a system with simulated data

                   2.  Beta Testing – testing a system with real data in the real world

                             a.  Recovery Testing – for the system to fail to verify that recovery operates properly

                             b.  Security Testing – verify that protection mechanisms work properly

                             c.  Stress Testing – try to break the system; what happens when the unexpected occurs?

                             d.  Performance Testing – determine how the system operates across a range of environments or conditions; often looking for response times or other performance measurements

 

III.  Installation – the process of changing over form the current system to the new one;  ***SEE TABLE 10-4, p. 375***

                   1.  Direct Cutover – completely remove old system and install entire new system at one time; fast, inexpensive (if done well), but the most dangerous method

                   2.  Phased (Modular) Cutover – install the new system one part (module) at a time, test it with the users, and when they are happy install the next component; longest method (time consuming), but works well with the users (they approve each module, one at a time)

                   3.  Pilot Cutover – install the new system for a select group of users, and when they are satisfied, install the system for everyone; this works well IF you choose the select group carefully

                   4.  Parallel Cutover – run both old and new systems in parallel for a time to be sure the new system’s outputs are correct, and the old system is still there if it’s needed; this is the safest method, but the most expensive (need two full sets of resources to run both systems all the time)

          Planning Installation – installations need to be planned, not just done in the course of the SDLC;

***** 5P’s:  Prior Planning Prevents Poor Performance *****

 

IV.  Documenting the System

          A.  User Documentation – users need to have an explanation on hand to learn and understand how to use the system

                   1.  System Documentation – specifications about the system

                   2.  Internal Documentation – system documentation that is part of the source code

                   3.  External Documentation – includes the outcome of all the structured diagramming for the system

                   4.  User Documentation – written or visual information specifically prepared for users

          B.  Preparing User Documentation

 

V.  Training and Supporting Users

          A.  Training Information System Users – users need to be trained in the workings and use of a new system

          B.  Supporting Information System Users – users will need ongoing education an problem-solving assistance

                   1.  Automating Support - online

                   2.  Providing Support through a Help Desk – a contact point with a real person

          C.  Support issues for the Analyst to Consider – recover and backup, disaster recovery, PC maintenance

 

VI.  Why Implementation Sometimes Fails – sometimes a system just does not work (or work well), and sometimes user do not like the system in spite of the fact that it works (low user satisfaction) – Stuff Happens!  That’s life, get used to it!

          A.  Risk – User Involvement can lower risk, but it adds to time and costs

          B.  Commitment to the project – the problem should be well understood and the solution should be approved at high levels in the firm – Top Management Commitment!

          C.  Commitment to change – are persons involved willing and able to change behaviors, procedures, etc?  Organizational (Cultural) change can ONLY start at the top and must be personally supported at that level

          D.  Extensive project definition and planning – the more planning that is done, the less likely failure

          E.  Realistic user expectations – Users must be managed!  The more realistic their expectations early in the project, the more satisfaction they will have at the end

 

VII.  Project Closedown – project review, team member reviews, close the contract (customer must “sign off”), a system audit (maybe)

 

VIII.  Conducting Systems Maintenance – fixes and enhancements to the system

          A.  Types of Maintenance

                   1.  Corrective Maintenance – fix mistakes

                   2.  Adaptive Maintenance – change functionality or evolve to changing business needs (environment)

                   3.  Perfective Maintenance – add new features or improve performance

                   4.  Preventive Maintenance – avoid future problems (prepare for Y2K)

          B.  The Cost of Maintenance – VERY Significant Cost!!!! Up from 50% 9years ago) to 80% in some organizations (aging systems, too expensive to replace, cost cuts); Environmental Changes – more and faster; get used to it!

          1.  Latent defects – unknown errors in the system; most maintenance costs are here (corrective maintenance)

          2.  Number of customers for a system – the more users, the more maintenance and training

          3.  Quality of system documentation – good documentation makes maintenance easier

          4.  Maintenance Personnel – maintenance is more difficult than development (and not as much fun)

          5.  Tools – CASE tools that can generate code, documentation, etc. can save a lot of time and costs

          6.  Well-structured programs – well-designed programs are easier to understand, fix, and enhance

          C.  Measuring Maintenance Effectiveness

                   1.  Number of failures

                   2.  Time between failures

                   3.  Type of failure

          D.  Controlling Maintenance Requests – which requests are met and which are declined?  Which are critical and which are wishful thinking?

                   Errors MUST be fixed!  Some are more critical than others

          E.  Configuration Management – the process of assuring that only authorized changes are made to the system

                   1.  Baseline Modules – software modules that have been tested and approved to be included in the most recent version of the system

                   2.  System Librarian – person responsible for checking out and checking in baseline modules (to approved persons for modification, and from persons who made approved changes; also making sure the code has been tested and passed the tests before check-in)

                   3.  Build Routines – guidelines for constructing the system

          F.  Role of CASE and Automated Development Tools in Maintenance – in a nutshell:  to speed development, reduce errors, and reduce costs

          G.  Web Site Maintenance

 

IX.  Maintaining an Information Systems at Pine Valley Furniture’s

 

X.  Electronic Commerce Application:  Systems Implementation and Operation

          Systems Implementation and Operation for Pine Valley Furniture’s WebStore

                   1.  Developing Test Cases for the WebStore

                   2.  Bug Tracking and System Evolution

                   3.  Alpha and Beta Testing the WebStore