We have a very large number of failed jobs(50k+) that I would like to remove but I cannot find a way to do this from cockpit. Is there some method of doing this that won’t cause database issues?
If you’re on the enterprise edition, you can create a batch operation from the menu. There, select delete running process instances as the action and filter the instances to those that have incidents (I assume that’s what you mean by failed jobs) and any other criteria you need to find the right ones.
With community edition, you would probably be best off creating a simple process that runs, selects a chunk of instances to delete, deletes them, then loops back to find the next chunk, etc, until there are no more candidates.
I have attempted this and have not had success. When I attempt a filter, none of the jobs come up. Is it possible that it is because it is from an earlier version of the pod? Would removing the earlier version remove the incidents? For clarification, I can only see the failed jobs when I click on the flow, and there’s a large number at the first step(50k) and I click that, and they appear at the bottom of the screen. There’s literally 50k of these failed jobs so I need to get rid of them.
Since I have direct access to the db, would there be any problem removing the records from the ACT_HI_INCIDENT table or are they also some place else?
Oh, on a side note, I don’t actually see the jobs in the batch window. I can only see them clicking on the incidents at the cockpit.
You need to filter first in the batch operation screen.
An earlier version of the pod should not influence the results in Cockpit; it goes into the database independent of your running application and shows you the current state.
Just removing the incidents in the database doesn’t really solve the underlying issue, so the process(es) will remain stuck there without proper resolution.
1 Like
I can’t seem to get them in the batch window no matter what I do.