An example of handling recovery in an end-to-end environment

If an operation has a conditional predecessor to be run on a fault-tolerant workstation, the scheduler takes into account any RECOVERY statement defined in the SCRPTLIB member related to this predecessor. If there is a RECOVERY statement defined, the scheduler postpones the evaluation of the condition according to the following logic:

• If the conditional dependency check is based on the status value C (Completed) or E (Ended-in-error):
  • If the recovery option is STOP or CONTINUE, the scheduler evaluates the condition when the recovery job, specified through the JOBCMD or JOBSCR parameter, ends. If there is no recovery job specified, the scheduler evaluates the condition after the reply to the prompt.
  • If the recovery option is RERUN, the scheduler evaluates the condition only if the rerun succeeds.
• If the conditional dependency check is based on the status value S (Started), the scheduler evaluates the condition immediately. If the recovery option is RERUN, the condition status is not reset after the rerun.