In case of System crash, Database errors or unorderly shutdown JobScheduler should update database with status and log for running tasks/orders

As of Behavior

1. If JobScheduler server crashes or admin unorderly shuts down the JobScheduler i.e. kill -9 <JS_PID> then JobScheduler can not save the status and log of running tasks/orders into the database.
2. After JobScheduler restart JOC shows empty values for Ended, Duration and Cause, since the end date field in the database is empty.

Expected behavior / Suggestions

• if JobScheduler service itself is being killed or server is crashed or database is not operating normally then realistically it is not possible for JobScheduler to save the status of running tasks/orders into database.
• After restart JobScheduler should update the database with as much as information possible for all the incompleted tasks/orders.
• JobScheduler can update the log field in the database with log files available from the file system.
• JobScheduler can frequently update the task/order status with status code "running", thus in case of above mentioned situation at least jobs can be marked as incomplete.

Maintainer Notes

• We tend to dismiss this feature request as it is targeted at providing higher availability with insufficient means. If high availability is required then a clustered database is a must. Any attempt to make JobScheduler compensate a missing database will result in limited success.
• The fact that in case of a missing database JobScheduler will continue and repeatedly try to connect to the database should not be misunderstood in a way that in this situation additional jobs could be started and the status of running jobs could be collected for later import into a database.