Celeryd Worker stops processing tasks

Question

I have a celeryd running with two workers, and watching them I see them accept 4 tasks each, process them, and then stop processing tasks. (Note that these tasks are long running, up to 2mins to process).

celeryctl provides the following information:

django@server: ./manage.py celeryctl inspect active
<- active
DEBUG 2012-06-01 12:51:11,330 amqplib 661 Start from server, version: 8.0, properties: {u'information': u'Licensed under the MPL.  See http://www.rabbitmq.com/', u'product': u'RabbitMQ', u'copyright': u'Copyright (C) 2007-2012 VMware, Inc.', u'capabilities': {}, u'platform': u'Erlang/OTP', u'version': u'2.8.2'}, mechanisms: [u'PLAIN', u'AMQPLAIN'], locales: [u'en_US']
DEBUG 2012-06-01 12:51:11,331 amqplib 507 Open OK! known_hosts []
DEBUG 2012-06-01 12:51:11,331 amqplib 70 using channel_id: 1
DEBUG 2012-06-01 12:51:11,332 amqplib 484 Channel open
-> eso-dev: OK
    - empty -

django@server:  ./manage.py celeryctl inspect scheduled
<- scheduled
DEBUG 2012-06-01 12:52:07,107 amqplib 661 Start from server, version: 8.0, properties: {u'information': u'Licensed under the MPL.  See http://www.rabbitmq.com/', u'product': u'RabbitMQ', u'copyright': u'Copyright (C) 2007-2012 VMware, Inc.', u'capabilities': {}, u'platform': u'Erlang/OTP', u'version': u'2.8.2'}, mechanisms: [u'PLAIN', u'AMQPLAIN'], locales: [u'en_US']
DEBUG 2012-06-01 12:52:07,108 amqplib 507 Open OK! known_hosts []
DEBUG 2012-06-01 12:52:07,108 amqplib 70 using channel_id: 1
DEBUG 2012-06-01 12:52:07,109 amqplib 484 Channel open
-> eso-dev: OK
    - empty -

django@server: ./manage.py celeryctl inspect registered
<- registered
DEBUG 2012-06-01 12:52:20,567 amqplib 661 Start from server, version: 8.0, properties: {u'information': u'Licensed under the MPL.  See http://www.rabbitmq.com/', u'product': u'RabbitMQ', u'copyright': u'Copyright (C) 2007-2012 VMware, Inc.', u'capabilities': {}, u'platform': u'Erlang/OTP', u'version': u'2.8.2'}, mechanisms: [u'PLAIN', u'AMQPLAIN'], locales: [u'en_US']
DEBUG 2012-06-01 12:52:20,568 amqplib 507 Open OK! known_hosts []
DEBUG 2012-06-01 12:52:20,568 amqplib 70 using channel_id: 1
DEBUG 2012-06-01 12:52:20,569 amqplib 484 Channel open
-> eso-dev: OK
    * celery.backend_cleanup
    * celery.chord
    * celery.chord_unlock
    *
celery_haystack.tasks.CeleryHaystackSignalHandler
    *
celery_haystack.tasks.CeleryHaystackUpdateIndex
    * convert.tasks.create_pdf
    * convert.tasks.create_pngs

In addition every time this happens the last thing to be printed to logs is:

[2012-06-01 12:17:53,777: INFO/MainProcess] Task convert.tasks.create_pdf[319984de-5bc4-47fc-891f-273d827d625f] retry: None
None
[2012-06-01 12:17:54,327: INFO/MainProcess] Task convert.tasks.create_pdf[8a89f3c1-e991-487e-a2db-a57d23bae17f] retry: None
None

The tasks also happen to have failed just before this is printed, and in my code all I have called is :

except HTTPError, e:
    statsd.incr('A.stat')
    log.warn('Woops: %s', e)
    create_pdf.retry()

If I kill celeryd (^C, and it dies straight away, no waiting for tasks) and start it again it continues like nothing has happend for a few more tasks and dies again (I think its always on the create_pdf task, but the logs show this task failing and being retried without a problem)

Answer 1

This is a know bug with celery that was fixed in v3.0. (Issue 707)