Measuring the baseline
Performance problems are often reported immediately following some change to system hardware or software. Unless there is a pre-change baseline measurement with which to compare post-change performance, quantification of the problem is impossible.
Changes to any of the following can affect performance:
- Hardware configuration - Adding, removing, or changing configurations such as how the disks are connected
- Operating system - Installing or updating a fileset, installing PTFs, and changing parameters
- Applications - Installing new versions and fixes
- Applications - Configuring or changing data placement
- Application tuning
- Tuning options in the operating system, RDBMS or an application
- Any changes
The best option is to measure the environment before and after each change. The alternative is running the measurements at regular intervals (for example, once a month) and save the output. When a problem is found, the previous capture can be used for comparison. It is worth collecting a series of outputs in order to support the diagnosis of a possible performance problem.
To maximize performance diagnosis, collect data for various periods of the working day, week, or month when performance is likely to be an issue. For example, you might have workload peaks as follows:
- In the middle of the mornings for online users
- During a late-night batch run
- During the end-of-month processing
- During major data loads