Jump to content


Swarm worker queue hangs

swarm

  • Please log in to reply
2 replies to this topic

#1 briand

briand

    Advanced Member

  • Members
  • PipPipPip
  • 77 posts

Posted 05 October 2016 - 11:22 PM

Both yesterday and today, we've had our worker queue in Swarm stop processing requests. We're running Swarm 2016.1 and Perforce server version 2012.1. We've had Swarm up and running on 2016.1 since May without any issues. The last config file change was in August, to update the email server. I have configured the max number of workers to five.

When I discovered that the queue processing has stopped (this was a couple of hours after the earliest queue entry still present), I performed an "apachectl stop" followed by "apachectl start". This unblocked the backup and the queue resumed processing.

Any thoughts regarding why this is happening? I don't see any errors or unusual messages in any of the logs I've looked at (I've looked at the Swarm log, the Apache logs, and the system log).

Thanks.
--
Brian

#2 briand

briand

    Advanced Member

  • Members
  • PipPipPip
  • 77 posts

Posted 06 October 2016 - 12:53 AM

It just happened again. I was monitoring for it, so I caught it quickly. I let seven minutes pass before I took action.

This time, I removed the lock files for all five workers (see: https://www.perforce...workers.restart ). The workers seemed to exit just fine (based on the queue/status output from my Swarm server, workers showed zero). Then I restarted the workers and the queue cleared out.
--
Brian

#3 briand

briand

    Advanced Member

  • Members
  • PipPipPip
  • 77 posts

Posted 07 October 2016 - 12:59 AM

Well, it looks like it is really a p4d issue. I'll start looking into my server side issues instead of Swarm.
--
Brian



Also tagged with one or more of these keywords: swarm

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users