[Core] Scheduler does not pay respect to running jobs

[peshkira]

The scheduler does not check if a job that is scheduled is still running and just runs another job. This can result in two or more jobs that reuse the same adaptor object, which potentially messes up the state of the adaptor.

When a new job should be scheduled, check if there is still another one for the same adaptor that is still running from the last time. If there is skip this schedule and wait for the next period. Repeat this check in every period.

Here is what happens if you schedule a pronom adaptor that takes longer than a minute to get the values

15:18:13,664 INFO  [stdout] (CoreScheduler_Worker-1) 15:18:13,664 INFO  [PronomAdaptor] Getting next batch of '200' with offset '800'
15:18:13,686 INFO  [stdout] (CoreScheduler_Worker-2) 15:18:13,686 INFO  [PronomAdaptor] Getting next batch of '200' with offset '800

[kduretec]

@peshkira please check and close if solved

[peshkira]

Commit 6cabb5f solves the issue (Thanks!) but has revealed another problem.
If some error occurs during data storage or somewhere else in the core, then the adaptor execution stops for some reason, but the scheduler continues to skip the new adaptor execution, because the exception event is not handled properly.

You can close this issue, but please create a new one for the other problem, or close this one, when the other problem is solved as well.

[luis100]

#162