When MIMIX is configured to replicate both job queue objects (*JOBQ) and jobs on those job queues, replicated jobs are placed in a staging area on the target system and are not actually submitted by MIMIX on the target system. The command string of the submitted job that has been replicated to the staging area is available on the target system until either MIMIX has processed a job completion journal entry for the source job or user action is taken to manually run or remove the job. This capability is useful in the event of a switch. The jobs from the original source system that did not complete before the switch can be submitted or discarded on the new source system after the switch. This can be done by steps within switch procedures for an application group or manually.
The replicated jobs that exist on a local system can be displayed using the Work with Replicated Jobs display (WRKREPJOB command) from the native user interface. During normal replication activity when this command is used from the target system, the listed jobs are automatically removed when MIMIX processes the journal entry for the job completion from the source system.
Following a planned or unplanned switch, the staging area on the new primary node may contain the command strings of replicated jobs that did not complete before the switch occurred. These jobs need to be resolved before another switch occurs. For application groups, you can enable optional steps within switch procedures that will run or remove any jobs in the staging area on the new primary node. For data groups that are not part of application groups, you must manually resolve the remaining jobs within the staging area by using options from the Work with Replicated Jobs display.