Improving distributed relational database performance through the system

Achieving efficient system performance requires a proper balance among system resources. Overusing any resource adversely affects performance. This topic describes the commands that are available to help you observe the performance of your system.

You can use the Performance Tools licensed program to help analyze the system performance. In addition, there are some system commands available to help you observe the performance of your system:

  • Work with System Status (WRKSYSSTS) command
  • Work with Disk Status (WRKDSKSTS) command
  • Work with Active Jobs (WRKACTJOB) command

In using them, you should observe system performance during typical levels of activity. For example, statistics gathered when no jobs are running on the system are of little value in assessing system performance. To observe the system performance, complete the following steps:

  1. Enter the WRKSYSSTS, WRKDSKSTS, or WRKACTJOB command.
  2. Allow the system to collect data for a minimum of 5 minutes.
  3. Press F5 (Refresh) to refresh the display and present the performance data.
  4. Tune your system based on the new data.

Press F10 (Restart) to restart the elapsed time counter.

Use both the Work with System Status (WRKSYSSTS) and the Work with Active Jobs (WRKACTJOB) commands when observing the performance of your system. With each observation period, you should examine and evaluate the measures of server performance against the goals you have set.

Some of the typical measures include:

  • Interactive throughput and response time, available from the WRKACTJOB display.
  • Batch throughput. Observe the AuxIO and CPU% values for active batch jobs.
  • Spool throughput. Observe the AuxIO and CPU% values for active writers.

Each time you make tuning adjustments, you should measure and compare all of your main performance measures. Make and evaluate adjustments one at a time.