An APM user running Asset Data Loads found that Data Load Processing Jobs are each assigned to a specific open Job Processing service when processing starts. If the process fails because the service goes down it remains assigned to the same service it was originally assigned to.
Users may spool up additional services during large data loads, etc, so which services are running at any given time varies. So if the original assigned job remains down or unavailable the retried Processing Job will not process. When the processing job fails and is retried it should once again look for an open service and reassign itself to the open service, not just the automatically assigned to the service that it was originally assigned to as it may not be running.
This was reported as a bug, ADO#11203446.
When the processing job fails and is retried it should once again look for an open service and reassign itself to the open service, not just the automatically assigned to the service that it was originally assigned to as it may not be running.
As a workaround it is suggested to add the "Server" column to the process job configuration and then search for the error processing job and use F2(edit in table) to manually overrode the assigned service to null/blank.
This forces the processing job to be reassigned at next run to another service.