Advanced Topics › Performance Considerations › Tuning Control File Access Rates › Control File Tuning: Analysis › Analysis Phase Summary

Analysis Phase Summary

By examining the DISPLAY IO and DISPLAY SERVICE statistics for every system in the MIMplex, you should be able to identify certain systems that are not performing control file cycles often enough to keep pace with your workload-intense systems. Because enqueue workload is the primary factor in driving control file cycles, CA MIM on systems with low enqueue workloads do not perform control file cycles as often as CA MIM tasks on workload-intense systems. CA MIM on low workload systems may access the control file only one time per second, while CA MIM on workload-intense systems may perform control file cycles 25 times per second.

Because of the disproportionate cycle rates (access frequencies), CA MIM on workload-intense systems will have to constantly reread and rewrite transactions destined for systems that are not accessing the control file often enough. The process of continuously rereading an rewriting latent transactions causes more control file blocks to be used, more control file I/O operations, elongated control file cycle times, inefficient transaction processing rates, and less than optimum throughput.

Based on the data in the DISPLAY IO and DISPLAY SERVICE illustrations used earlier in this section, the following analysis can be made: the MIMplex is communicating across systems using a DASD control file. It is performing control file cycles at a rate of roughly 22 times per second. Control file cycles are most likely driven by the fact that it has a fairly high enqueue workload (the number of requests is 16,896,706). This MIMplex appears to be poorly tuned because of the following facts:

• The number of blocks read and written are more than the acceptable value of 1.1 for COMMUNICATION=DASDONLY.
• The average control file cycle time is longer than the acceptable value of 80ms (for 3390 DASD control file).
• The transactions read to transactions processed ratio is more than the acceptable 10:1.
• The enqueue time per request is above the highest acceptable value of 70ms. This is the most important factor.

In short, there seems to be one or more external systems in the MIMplex that are not accessing the control file often enough, and this system is constantly having to reread and rewrite the transactions of that system. The DISPLAY IO and DISPLAY SERVICE data from the external systems should be examined to determine which system is not accessing the control file often enough.

Note: To see how this can be corrected, see Control File Tuning: Example in this chapter.

The purpose of the analysis phase is to help you identify which systems may not be accessing the control file often enough in your MIMplex. It can also be used to identify which systems may be dominating the control file due to enormous enqueue workloads, although this is rare. In any case, CA MIM parameters can be defined to increase or decrease control file cycles if the analysis phase shows CA MIM on certain systems to be accessing the control file too often, or not enough. The control file tuning sections on parameters and implementation in this chapter explain how to use these CA MIM parameters to increase or decrease control file cycles on specific systems.