Jobs Skipped When Scheduler Restarts After Crash

Can you tell me how 24x7 is supposed to behave in regard to time-scheduled jobs whose time is past when the scheduler restarts after a crash?

Here's the scenario. Job "AR" is scheduled for 1:27am. The 24x7 service crashes on some kind of internal error at 1:25am (this is a separate problem--I can send you the event log messages).

A monitoring application (in-house application) see that 24x7 has crashed and gets the service started again by 1:27:30am. The job "AR" never runs.

So when restarted after a crash, should jobs whose time has gone by run? Or should they be skipped? Can I do something to force them to always be run?

For the job AR to start after restart, the "reset job queues on startup" system option must be unchecked. The scheduler will then reload job queues as there were before the crash.

Now, in order to avoid dealing with such issues, please make sure all jobs are set to run detached and synchronous. In a very rare circumstances there could be exceptions to this rule and jobs set to run non-detached. If needed, create 2 or 3 additional jobs queues to paralelize job sterams.

Note, detached jobs run in a separate memory space and so they cannot crash the scheduler as well as their memory leaks and other problems do not affect the memory of the scheduler process.