For some installations, a measurement and evaluation plan might be suitable.
A measurement and evaluation plan is a structured way to measure, evaluate,
and monitor the system’s performance. By taking part in setting up this
plan, the users, user management, and your own management will know how the
system’s performance is to be measured. In addition, you will be able
to incorporate some of their ideas and tools, and they will be able to understand
and concur with the plan, support you and feel part of the process, and provide
you with feedback.
The implementation steps for this plan are:
- Devise the plan
- Review the plan
- Implement the plan
- Revise and upgrade the plan as necessary.
Major activities in using the plan are:
- Collect information periodically to determine:
- Whether objectives have been met
- Transaction activity
- Resource utilization.
- Summarize and analyze the information. For this activity:
- Plot volumes and averages on a chart at a specified frequency
- Plot resource utilization on a chart at a specified frequency
- Log unusual conditions on a daily log
- Review the logs and charts weekly.
- Make or recommend changes if objectives have not been met.
- Relate past, current, and projected:
- Transaction activity
- Resource utilization.
to determine:
- If objectives continue to be met
- When resources are being used beyond an efficient capacity.
- Keep interested parties informed by means of informal reports, written
reports, and monthly meetings.
A typical measurement and evaluation plan might include the following items
as objectives, with statements of recording frequency and the measurement
tool to be used:
- Volume and response time for each department
- Network activity:
- Total transactions
- Tasks per second
- Total by transaction type
- Hourly transaction volume (total, and by transaction).
- Resource utilization examples:
- DSA utilization
- Processor utilization with CICS®
- Paging rate for CICS and for the system
- Channel utilization
- Device utilization
- Data set utilization
- Line utilization.
- Unusual conditions:
- Network problems
- Application problems
- Operator problems
- Transaction count for entry to transaction classes
- SOS occurrences
- Storage violations
- Device problems (not associated with the communications network)
- System outage
- CICS outage time.
[[ Contents Previous Page | Next Page Index ]]