Restarting a non-unique job step

If a job step that is defined as unique abends and requires restart, the usual procedure is to correct the problem that caused the abend and submit the job step for restart. No application program or JCL changes are required. However, if a job step that is defined as non-unique abends and requires restart, you must perform the following steps:
  1. Correct the problem that caused the abend.
  2. Obtain the restart data set name from the restart control record for the abended job step. The Reference section describes how to access this record with action code S. The JES job number may help you identify a non-unique job step; you can display this job number by selecting the restart control record.
  3. Include the ARCXRST DD statement, defining the data set name you obtained in the previous step, in the restart JCL.
  4. Submit the job step for restart.


    If the job step is erroneously defined as non-unique, perform steps 1 through 4, wait for the restart to complete successfully, and then change the value of the Non-Unique Job Step option. For more information, see Reference.

This version of the documentation is no longer supported. However, the documentation is available for your convenience. You will not be able to leave comments.