The job journal is a sequential data set that resides on the spool volume of the job entry subsystem. It contains a set of selected job-related control blocks that are critical to automatic restart processing.

The job journal is necessary because the scheduler control blocks are maintained in the scheduler work area (SWA) in pageable storage, rather than in a job queue on external storage. When a job or the system itself fails, the space that contains the SWA and its job control blocks is lost. Because it preserves up-to-date copies of certain critical control blocks, the job journal makes it possible to reconstruct the SWA. SWA control blocks are restored to the state they were in just prior to the failing step for automatic step restart. For automatic checkpoint/restart, SWA control blocks are reconstructed as they appeared at the time of the most recently issued CHKPT.

Job journaling is required before a job is eligible for an automatic restart.

The system creates a job journal to hold restart information for any job specifying job journaling. After a system failure and the system restart of the failing main processor, the jobs in execution that requested job journaling are restarted (warm started) by the system. If a job is eligible for automatic restart, the operator is sent the message IEF225D asking if the job should be restarted. If the job is not eligible for restart or if the operator indicates that the restart should not be attempted, any scratch or VIO data sets the job allocated are deleted, and the job terminates. Jobs that frequently use scratch or VIO data sets should request job journaling.