Restarting synchronized jobs

The macro separates the original JCL member into n + 1 members using the default prefix of JOB for the member names. Thus, JOB000 is the restart job used to reset the synchronization file and monitor job completion, and JOB001 - JOB00n are the separate recovery jobs generated by BMC AMI Recovery Manager.

If you choose to restart the jobs, copy the ARMSBGEN member from the .DBCLIB library to a library in your SYSPROC concatenation, then perform the following steps:

To restart a synchronized job

1. For each recover job (1 - n), locate the job output and note the step name in which the job got the first return code greater than 4. Some jobs may have completed successfully and will not need to be rerun.

2. If the original JCL is not in a PDS, copy it to a PDS and execute ARMSBGEN to separate the JCL into its component jobs.

   • Edit the BMC AMI Recovery Manager-generated member using ISPF EDIT.
   • Type ARMSBGEN prefix on the command line, where prefix is an optional 1 to 5 character prefix to use for the member name. If prefix is not specified, ARMSBGEN creates members with a prefix of JOB.

   Important

   After ARMSBGEN is complete, it CANCELs out of the ISPF EDIT, leaving the BMC AMI Recovery Manager-generated JCL intact.

3.  Edit member JOB000 (or prefix000). ARMSBGEN has added the following card to the JCL directly after the job card:

   // RESTART=ARM0002.ARMSYNC

   Perform the following substeps:

   1. Place a comma at the end of your job card and verify the restart card is immediately after your job card. Do not change the step name on the RESTART card. The job is now ready to prepare the synchronization file for restart and clean up again at the end as needed.
   2. Submit this job before you submit other recover jobs. It waits on the other jobs to complete before it completes.

4.  Edit member JOB001. ARMSBGEN has added the following card to the JCL directly after the job card:

   // RESTART=ARM?????

   Perform the following substeps:

   1. Place a comma at the end of your job card and ensure the restart card is located immediately after your job card.
   2. Change the ARM????? on the restart card to the step name of the first step that received a return code greater than 4, if any, in the original execution.

   3. If the failing step was DSNUTILB, you may need to add RESTART to the parameter list on the DSNUTILB execution.For example, if the subsystem ID is DBDF, the parameter list should look like this:

      EXEC PGM=DSNUTILB,PARM='DBDF,,RESTART'

      Important

      Adding the RESTART parameter is required only if DSNUTILB was started and placed in a STOPPED status at the time of the failure.

5. Repeat Step 4 for each remaining job (JOB002 - JOB00n).
6. Submit member JOB000 if you have not already done so (see Step 3).

7. Submit members JOB001 - JOB00 n that require restart due to previous failures.

   Important

   If for any reason you need to restart again, change the restart parameters as required in JOB001 - JOB00 n and resubmit JOB000 - JOB00 n.