#celery

hi, anyone have experience trying to use subprocess within a celery task

Hi, ... is there a way to get some result or notification back in case the invoked task is not found in the worker?

as for now the task sender only sees an asyncresult in state pending ... but task message was already consumed by a worker who doesn't know the task. and the sending app doesn't get any information that the job couldn't be started

is this expected behaviour?

or maybe the other way round ... is there a way to find out whether a job has been rejected or not?

hi there

i'm trying to use retry() and it is working fine with CELERY_ALWAYS_EAGER=True but with CELERY_ALWAYS_EAGER=False I am seemingly running into the problem where the request context is empty

it's described here: https://groups.google.com/forum/#!topic/celery-...

but I am on 3.1.11 and am using bind=True and self.retry

so I'm at a bit of a loss

.wub 2

eedeep: Calling the task on the worker or calling the task directly (task.__call__) ?

do I have to copy tasks definition in every server in order to be able to distribute celery over many servers?

asksol: hi there sorry I had to ride home

asksol: calling the task via the_task.delay()

with ALWAYS_EAGER it works perfectly

but run async proper, by the worker, it actually does retry once but I can see via logging statements that the request context for the resubmitted task is empty

which looks to me exactly like what is described here https://groups.google.com/forum/#!topic/celery-...

which you provide a solution for...but I'm definitely doing what you say there, but seemingly to no avail

eedeep: what celery version? are you using windows?

asksol: thanks for your thousands of replies to issues and forum posts btw...you're everywhere that celery is

asksol: 3.1.11 on ubuntu 12.04 with rabbitmq 2.7.1

asksol: this is the task https://gist.github.com/eedeep/bee8b906e6ffd349...

it's messy as there cos I've been messing around trying to figure out what's going wrong