After the system reads and interprets the restart job, it reads the specified checkpoint entry and merges information from it into the scheduler work area entry for the job. As a result, the work queue entry differs from the entry existing during the original execution, as described earlier. See How the Job Is Reinterpreted.

Next, the system initiates the restart step normally. The system reads the specified checkpoint entry again and functions as in automatic restart. Restart is delayed until the required virtual storage area is available.

When a job restarts correctly, the system operator receives two messages: IHJ006I and IHJ008I. If these messages do not appear for nonpageable storage jobs (ADDRSPC=REAL), enter a DISPLAY A command to see if a system task or another job is using the required storage. The system operator can then stop or cancel the conflicting job.

If there are multiple tape or disk volumes, the system may ask the operator to mount data volumes other than those required at the beginning of the job. In addition, any card input data sets that are used by the failing job step must again be made available to the system.