Details
-
Bug
-
Status: Open
-
Major
-
Resolution: Unresolved
-
None
-
None
-
None
-
None
Description
currently it starts the JobCoordinator server, starts the job, and stops the JobCoordinator server right away. There is not need for a separate server here, because the job model is passed directly to the container. And if container dies it is not going to be restarted anyway.
It seems that we can just disable both start and stop of JobCoordinator in ThreadJobCoordinator.