Aborted jobs in celery broker remain stuck on

[glebkuznetsov]

Nothing is running but I'm seeing a celery.chord_unlock() message being repeated every 1 sec in the logs.

I'm not sure what the immediate repercussions of this are besides filling up the log file endlessly.

We somehow need a way to detect and clear the celery queue when there are no real jobs. How can we detect this?

[glebkuznetsov]

Just noting the way to clear these stuck tasks:

1. Stop celery, e.g.

   supervisorctl stop celery

2. From manage.py shell or in code:

   from celery.task.control import discard_all
   discard_all()

3. Restart celery. should be clear:

   supervisorctl start celery