| Issue Number | 4235 |
|---|---|
| Summary | Publishing Doppelgänger? |
| Created | 2017-03-06 14:16:15 |
| Issue Type | Bug |
| Submitted By | Englisch, Volker (NIH/NCI) [C] |
| Assigned To | Englisch, Volker (NIH/NCI) [C] |
| Status | Closed |
| Resolved | 2017-03-14 11:20:46 |
| Resolution | Cannot Reproduce |
| Path | /home/bkline/backups/jira/ocecdr/issue.204157 |
In the past I've noticed when I'm starting a scheduled job by
clicking the "Custom Run" button that a second job is being submitted. I
originally thought I may have clicked the "Submit" button twice but I am
certain this was not the case today.
I've submitted a weekly publishing job at 13:58h and at 14:04h the same
job was submitted again. Obviously, the second job fails with the error
message:
Failure - JobNNNNN still in process
We need to look into what's submitting this second job.
Which tier?
So far I've only tried it on DEV and QA. I don't remember if I tried it on STAGE, too.
I don't recall if we've even set up the CDR scheduler jobs on stage. Which tier did the incident you described at the top of this ticket happen on?
I did setup some jobs on STAGE.
The incident that triggered me to enter the ticket today happened on QA.
The nightly publishing job started on STAGE and it failed -
twice!
The job started at 18:00h and at 18:05h. It seems the fact that I tried
to submit a custom run of the publishing job had nothing to do with the
two job submissions.
I think it might have something to do with the fact that we have installed publishing jobs in the new CDR scheduler, without having disabled them in the Windows scheduler. The Windows scheduler has the nightly publishing job set up to run at 6:00 pm, and the CDR Scheduler has it set to run at 6:05 pm. We'll have to look at the publishing log to find out what the failures were all about.
"... too many errors detected at CDR37914."
The 18:00h and 18:05h failures were related to a former job still in the Verifying state. The latest failure you're mentioning is related to a new filter not yet installed.
I have to agree with Bob, this "Problem" was likely a result of
myself starting the publishing job a couple of times right around the
regularly scheduled starting time of the job and I had forgotten that
the jobs on STAGE had not yet been disabled in the Windows
Scheduler.
Since I've reported the issue I've started publishing jobs a few more
times using our scheduler and those jobs started and ran fine.
I'm closing this ticket.
Elapsed: 0:00:00.001284