1

The rate of abends in the data center can indicate how effective the data processing organization is in meeting its goals. However, you must determine how to distinguish production jobs from test jobs. Test jobs must be excluded for audit purposes.

2

Determine from the Job History Search display (4.4) if any key systems, such as IMS or CICS, have abended. You can usually identify key systems from the Subsystem Display (3.1).

Obtain job names from the Operations staff or the Search JCL Libraries display (4.3) if you know the library names of production JCL libraries.

3

Take a sample of recent production job history and determine from the display which jobs have abended.

Determine if adequate documentation is available for operators, which explains how to recover from abends.

4

Determine from the display if any 0Cx‑series abends for production jobs, such as 0C7 abends, exist. These are program checks that can indicate that program testing before promotion to production is insufficient.

5

Determine from the display if any B37, 80A, 213, or 913 abends for production jobs exist. These error conditions can indicate a problem in the method that the data center uses to control production jobs.

6

Determine from the display if any 122 or 222 abends for production jobs exist. These abends are the result of operator cancel commands. Frequent cancels can indicate operational problems.

Check also for A22 abends, as these result if the operator has issued the FORCE command to terminate an address space. Frequent FORCE usage might also indicate operational problems.

7

Determine from the display if any 322 or 722 abends for production jobs exist. These abends are the result of automatic cancel commands being issued for the job when it has exceeded installation-specific and job class-specific thresholds for CPU time consumption and JESx output lines. Frequent instances of either abend can indicate problems with the applications themselves (CPU/printer loops), or they might be normal and acceptable, in which case the job’s allowable resource limits would have to be reevaluated.

8

Note the start and completion times recorded for a sampling of jobs in the operator’s shift log or schedule.

9

Select a sample of the jobs identified in Step 7. Verify that the start and completion times recorded for the job in the shift log are the same times recorded for that job on the Job History Search display (4.4).