DjangoBot: #celery

aha, so closing the terminal sends the HUP signal

it will also effectively close stdin+stdout, making the restarted celery a daemon :)

I have no idea how to fix it so it terminates when you close the terminal window

additionaly there was a bug, because the worker processes had the same HUP handler as the main process, so when you close the window, each worker will start a new celeryd :/

i guess that's what SIGHUP is for, just seen a lof or programs use it to reload configuration

guess it could skip the HUP handler if celeryd is not running as a daemon

apache also uses HUP to reload configuration

asksol: hey, have you seen this error before? http://paste.mitechie.com/show/100/

I'm going to update the server to 2.0.1 and see if it changes but not seeing it in the changelog

currently at 2.0

no, never

maybe ask the sqlalchemy guys, no idea what that is

k, seems like some bug in sqlalchemy and cleaning up tasks?

I'm ignoring results so wondering if something is still trying to run cleanup

it doesn't seem related to cleaning up tasks

it's from retrievedata in /srv/venv/qmail_ve/qmail_app/qmail_app/celery_app/tasks.py

oh doh

nvm, mondays

finally fixed the problem when closing the terminal window running celeryd

and http://github.com/ask/celery/commit/d335775246a...

asksol: you seen this 'Missing hostname for AMQP connection.' before, despite the fact celeryd is connected....

celeryd and celeryev appear to work fine, but views are crashing out... trying to upgrade 1.0.3 to 2.0.1

matclayton: add os.environ["CELERY_LOADER"] = "django" to your .wsgi file

got that there

issue is in manage.py runserver

hmm, that should work

and you have djcelery in INSTALLED_APPS?

yes

in the views, try dumping out celery.loaders.setup_loader()

demoapp/views.py and runserver works there

hmmm we have some strange settings loaders

could be that, but I doubt it

try adding import djcelery directly in settings.py

ok, that also removed the No celeryconfig.py msg

testing it now

appears to have fixed it, I also added the database backend (changed from amqp)

just going to see which it was

yeah the import fixed it

nice. I added djcelery.setup_loader() so it seems more explicit

cool

seems django is lazily loading the models now then

might take celerymon out for a spin now

2.0.1 seems ok at first glance

as in its all working

interesting celerymon is 404ing, but least the server fires up

celerymon is *very* experimental

I don't have the front end coders to finish it yet

[No longer have to type PYTHONPATH=. to use celeryconfig in current dir] http://github.com/ask/celery/commit/c069f7d6a8a...

ok cool, was just playing with it

btw

there is no index

see the source code for list of urls

ah ok :)

Hi, How can I run periodic task every minute between 9 to 10:30, it works for 9 to 10 or 11!

crontab(hour="9,10", minute="0,30") ?

ask: Let me try this, thanks

asksol, sorry to bother you, but is there a way in 2.0.X to inspect scheduled tasks?

I think the doc here are out of date -- http://ask.github.com/celery/userguide/workers....

that is for the master branch

ahh, my mistake

oh, it's on celeryq.org as well

the command is broadcast("dump_schedule", reply=True)

celery.task.control.broadcast

asksol, thank you :)

but far more useful in the master branch, returns a lot better info

@ask: I think crontab(hour="9,10", minute="0,30") will execute at 9, 9:30, 10 and 10:30

for periodic tasks, is anything required other than decorating tasks in tasks.py with @periodic_task, running celerybeat, and running celeryd?

claymation: not really

I had forgotten to register my task :-/

run celerybeat with --loglevel=DEBUG for more info

thanks, though

<-- n00b

hello - just getting started with celery in django and having some trouble with using ghettoq

everything works fine with rabbitmq, but if i drop the BROKER_* settings and use CARROT_BACKEND = 'ghettoq.taproot.Database' instead, celery complains of a missing AMQP hostname

Hi all, I'm having a really difficult time debugging this issue: http://dpaste.com/219825/ I can't figure out from that traceback what object the __init__() method is being called on

savid: the exception is not pickleable

see http://github.com/sciyoshi/pyfacebook/issues#is... for a similar issue

asksol, interresting

asksol, would be nice to at least see the exception info ... does celery at least store the exception info somewhere?

ie, somewhere in plain-text?

what version is this?

thing is, it's an imroper exception. it tries to fix them up, but there are probably still cases where it can't

2.0.0 does fix a lot more of them though

does anyone use celeryd-multi?

i can't seem to get a celeryd worker started with ghettoq + sqlite

enqueing tasks is working now

asksol, looks like celery 1.0.2

asksol, any idea if there is any backward-incompatibility to from 1.0.2 to 2.0?