Documenting the monitoring run

For each trace interval there are several general reports or overall summaries of the processing that took place. You can use these reports as part of your IMS Monitor run documentation.

It is important to record, as accurately as possible the conditions under which the trace was taken. Your documentation can include system status information obtained by the /DISPLAY command several times before and after the trace, an expected profile of the application program activity, and any desired processing events. The trace interval should represent typical processing loads and not be a biased or inadequate historical record.

Adding to the System-Configuration report data

The first general report titled SYSTEM CONFIGURATION is found under the page heading BUFFER POOL STATISTICS. It shows the modification level of the IMS and z/OS® systems. The system configuration output is illustrated in the following section.

Recording the Monitor trace interval

The heading of most IMS Monitor reports carries the trace start and stop times. It is shown in the format YEAR DAY (Julian) HH:MM:SS. The overall length of the trace interval is given in milliseconds under the title MONITOR OVERHEAD DATA. The following line shows how many trace records were placed on the IMSMON data set. An example of the monitor trace interval recording is shown in the following example.

***IMS MONITOR***  BUFFER POOL STATISTICS   TRACE START 2009 180   5:55:15   TRACE STOP  2009 180   5:59:49  PAGE 0001
 
                   S Y S T E M   C O N F I G U R A T I O N
 
                                                                 SYSTEM CONFIGURATION       :  MVS/ESA
                                                                 IMS VERSION                :  11
                                                                 RELEASE LEVEL              :   1
                                                                 MODIFICATION NUMBER        :   0

Completing the Monitor run profile

A compact set of processing ratios is found at the end of the Run-Profile report. The statistics summarize, for the monitor interval, the transaction throughput and the degree of DL/I and I/O activity. An example of the run-profile report is shown in the following example.

The lower part of the Run-Profile report shows several ratios:

  • Program elapsed time to DL/I elapsed time for each region
  • DL/I elapsed time to wait time during DL/I processing
  • Program elapsed time to other subsystem call elapsed time
  • DL/I elapsed time to other subsystem call elapsed time

Each dependent region is identified by a sequence number, starting at region 1.

    IMS MONITOR   **RUN PROFILE**            TRACE START 2009 180   5:55:15     TRACE STOP  2009 180   5:59:49  PAGE 0184
TRACE ELAPSED TIME IN SECONDS.............274.6
TOTAL NUMBER OF MESSAGES DEQUEUED........1403
TOTAL NUMBER OF SCHEDULES.................173
NUMBER OF TRANSACTIONS PER SECOND...........5.1
TOTAL NUMBER OF DL/I CALLS ISSUED.......18632
NUMBER OF DL/I CALLS PER TRANSACTION.......13.2
NUMBER OF OSAM BUFFER POOL I/O'S........11236,    8.0 PER TRANSACTION
NUMBER OF MESSAGE QUEUE POOL I/O'S..........0,    0.0 PER TRANSACTION
NUMBER OF FORMAT BUFFER POOL I/O'S..........0,    0.0 PER TRANSACTION
RATIO OF PROGRAM ELAPSED TO DL/I ELAPSED:
              REGION   1:   1.09
              REGION   2:   1.09
              REGION   3:   1.00
              REGION   4:   1.02
              REGION   5:   1.01
              REGION   6:   1.00
              REGION   7:   1.00
              REGION   8:   1.00
              REGION   9:   1.17
              REGION  10:   1.00
              REGION  11:   1.00
              REGION  49:   1.03
              REGION  50:   1.19
RATIO OF DL/I ELAPSED TO DL/I IWAIT:
              REGION   1: 325.65
              REGION   2:  73.49
              REGION   4: 100.35
              REGION   5:  85.76
              REGION   6:  82.99
              REGION  47:  95.64
              REGION  48:  45.93
              REGION  49:   9.22

You can match the regions to the name of the z/OS job using the Region-and-Jobname report. The job names correspond to the step names on the EXEC statements of all the dependent regions started by the operator before the trace was started. The region job names are included on the monitor output page with the heading GENERAL REPORTS, as illustrated in Detecting checkpoint effects.

Some generalized processing ratios are given at the end of several buffer pool statistics reports. You can include them in the documented profile of the trace interval. These are not specific to one application or system resource but can be used as indicators of variation across a series of monitor runs.

The ratios are:

  • The total number of OSAM reads + OSAM writes + all waits divided by the total number of transactions.

    From the Message-Queue-Pool report, this ratio indicates on a per transaction basis the physical I/O activity required to handle the message queuing function.

  • The total number of OSAM reads + OSAM writes + BISAM reads divided by the total number of transactions.

    From the Database-Buffer-Pool report, this ratio indicates on a per transaction basis the physical I/O activity required to handle the database buffering function.

  • The total prefetch I/Os + immediate fetch I/Os + directory I/Os divided by the total number of transactions.

    From the Message Format Buffer Pool report, this ratio indicates on a per transaction basis the physical I/O activity required to handle the MFS function.