Is backburner:threads_on_fork:work reliable ? about backburner HOT 8 CLOSED

BTW I don't reproduce the problem with backburner:forking:work
But I don't know how to define how many processes should run in parallel for each job with the Fork strategy.

from backburner.

Is it linked to #47 ?

from backburner.

It may be. In investigating the connection deadlock issue in #47 I found that the beanstalkd DEADLINE_SOON message (120 seconds by default) came through, which causes the job to fail and retry. Therefore, if you didn't override the beanstalkd timeout, and your job fails and retries after 120 seconds, you're probably experiencing the same thing I was in that issue.

Another way to check would be to use the threads_on_fork worker with only one thread. If it reliably processes jobs, you're probably running into the aforementioned deadlock issue.

from backburner.

Thanks for your reply !
I will try with one thread. Hope your fix will get on the master branch soon !

from backburner.

Pushed 0.4.5 with @thcrock pull request. Thanks!

from backburner.

Yipee ! Gonna try it today and close the issue if it fixes it ! Thanks guys.

from backburner.

looks like it does ! thanks a lot 👍

from backburner.

Great thanks for the update.

from backburner.