#celery

hmm, so in a cluster, does celery communicate directly with each other or through something like rabbitmq?

MACscr: every celeryd worker talks to the broker, which may be rabbitmq, yes

After a task is done, and the result is stored in the result backend database.. is it possible to figure out what task originally triggered it?

I get a python error when doing a Group() of only one task. Is that expected?

matt1s, afaik not unless you store information on tasks somewhere else. or use custom id's

nanonyme: Hm okay, where can I read about custom IDs?

1 month to install broadband in the new apartment :(

would think I was living in some cave somewhere, but no this is London

matt1s: results only store a list of children, there's no link back to parent

could be something to add

asksol: I am using djcelery with the database backend for the results, not sure if you are familliar with it? It has a "meta" column that contains the exact same hash for two task instances of the same task.. Trying to figure out what that column actually contains. If not, I can store more information in the Django database

task_id should be unique, so if you have two rows with the same id something wrong happened

asksol: the task_ids are different, but each row has a "meta" column that contains the same hash (I think it is a hash).

asksol: reading through the source now to make sense of it.. Just hoping the column contains something that I can use to identify the originating task

yeah, that is arbitrary data

anything can be stored there

asksol: Thanks, then I can store it there!