eG Monitoring
 

Measures reported by AS400JobMntrTest

Sometimes, administrators may only want to monitor a specific set of jobs for the purpose of analyzing the critical performance of the target AS400 server and troubleshooting the issues such as resource contention. This purpose can be easily achieved using the AS400JobMntrTest test. This test lets administrators to set a pattern to filter out and monitor the jobs of their choice. The pattern can be defined using the USER NAME, JOB NAME, JOB TYPE and/or SUB SYSTEM parameters. To define the pattern, administrators can specify all of these parameters or anyone of these parameters or the combination of these parameters. This test will therefore only monitor the jobs that are filtered on the basis of the pattern defined by administrators.

Note:

To add the jobs for monitoring by this test, administrators should configure at least one of the USER NAME, JOB NAME, JOB TYPE and SUB SYSTEM parameters.

This test auto-discovers the jobs according to a specific pattern that has been configured by administrators. For each pattern, this test reports the following metrics:

  • the number of active jobs;

  • the jobs held in the queues;

  • the number of threads required to perform the jobs;

  • how long the CPU resources were utilized by the jobs;

  • the amount of temporary storage allocated for executing the jobs.

This way, administrators are alerted to slowness in processing the jobs, a processing bottleneck on the queues and the CPU resource contention created when the jobs use the resources for a prolonged time. Thus, administrators can initiate the remedial measures before the users complaint about any performance lag.

Outputs of the test : One set of results for every pattern configured by administrators.

The measures made by this test are as follows:

Measurement Description Measurement Unit Interpretation
No_of_active_jobs Indicates the number of active jobs that match this pattern. Number  
No_of_jobQ_jobs Indicates the number of jobs, that are submitted, but made to wait in a queue before they could begin running. Number A very high value of this measure is a cause for concern, and would require investigation.
No_of_outQ_jobs Indicates the number of jobs, held in the output queue, to be sent to a subsystem. Number  
Total_threads Indicates the number threads utilized to perform the jobs that match this pattern. Number Ideally, the value of this measure should be high. A significant decrease in the value of this measure shows that the jobs are being executed slowly. As a result, the system performance will be degraded and the user experience will be affected.
Total_cpu_used Indicates the CPU time used to perform the jobs that match this pattern. Secs Ideally, the value of this measure is preferred to be low. A high value indicates that the jobs are utilizing the CPU resources for longer duration. This may lead to CPU resource contention on the server.
Total_temporary_storage Indicates the amount of temporary storage allocated for the jobs that match this pattern. MB