hi looking for more info on celery.graph. The celery doc seem to have only api info... Any pointers
amcm joined the channel
maryokhin has quit
skbd has quit
italorossi joined the channel
italorossi has quit
italorossi joined the channel
italorossi has quit
italorossi joined the channel
cjhubert joined the channel
maryokhin joined the channel
linkedinyou joined the channel
ASUChander joined the channel
ASUChander
Hi, I have a django app(1.9) using Celery (3.1) and am having some issues - I can run the worker in debug mode and it shows it registers all my tasks properly, but then the worker exits with exit code 0 (no errors, etc.) any ideas as to why?
I'm using Kombu/db as the messaging queue
super low volume tasks
saurabhjha has quit
saurabhjha joined the channel
Debnet`Laptop joined the channel
joelcox joined the channel
taggers joined the channel
Debnet`Work has quit
the_rat_ joined the channel
maryokhin has quit
the_rat_ has quit
linkedinyou has quit
joelcox
I'm converting a home-grown message task system to Celery. In the original a message was published with a routing key "post.created", which was routed to two queues in a topic exchange "post-upload" and "post-process" which are both bound to the "post.created" routing key. How would I replicate this in Celery?
Basically I want to trigger two tasks at the same time :)
Debnet`Work joined the channel
saurabhjha has quit
Celery seems to have groups, but I'd prefer to define these kind of things in my routing..
or is this "unceleryic" :)
Debnet`Laptop has quit
nicksloan has left the channel
saurabhjha joined the channel
domino14 joined the channel
saurabhjha has quit
domino14_ joined the channel
saurabhjha joined the channel
saurabhjha has quit
gnoze5 joined the channel
domino14 has quit
maryokhin joined the channel
domino14_ has quit
linkedinyou joined the channel
gnoze5 has quit
italorossi has quit
joelcox has quit
maryokhin has quit
Tawonga has quit
elyssonmr has quit
Tawonga joined the channel
ASUChander
I'm running celery w/Django and when I run the worker in a shell, everything is fine - works great. However, when I launch it using "multi start MYAPP" I find that it uses all defaults (so it doesn't use the configured broker/creds) any ideas as to what the cause might be?
colinbits joined the channel
italorossi joined the channel
Roehmer joined the channel
italorossi has quit
Debnet`Work has quit
atomekk has quit
the_rat_ joined the channel
adrian_lc has quit
darkelda has quit
Iniglo joined the channel
the_rat_ has quit
domino14 joined the channel
linkedinyou has quit
nickabbey joined the channel
Iniglo has quit
nickabbe_ joined the channel
JamesGBI_ has quit
nickabbey has quit
JamesGBI joined the channel
JamesGBI has quit
maryokhin joined the channel
drag0niu1 joined the channel
drag0nius has quit
brambles joined the channel
domino14 joined the channel
cellopin_xx joined the channel
mrud_ joined the channel
Callek_ joined the channel
EnTeQuAk has quit
patrick91 has quit
magicflakes has quit
wahnf___ has quit
linovia has quit
syphar has quit
TenukiBen has quit
ASUChander has quit
phretor has quit
primozk has quit
mrueg has quit
d-fence has quit
joar joined the channel
phretor joined the channel
cellopin_xx is now known as magicflakes
joar has quit
joar joined the channel
italorossi joined the channel
mrueg joined the channel
joar joined the channel
joar joined the channel
avium joined the channel
primozk joined the channel
minsky joined the channel
d-fence joined the channel
italorossi has quit
dhanush joined the channel
patrick91 joined the channel
EnTeQuAk joined the channel
lkannan joined the channel
linovia joined the channel
merlinsbrain joined the channel
akitada joined the channel
syphar joined the channel
italorossi joined the channel
wahnf___ joined the channel
TenukiBen joined the channel
bgyss joined the channel
elyssonmr joined the channel
malinoff joined the channel
italorossi has quit
Ergo joined the channel
italorossi joined the channel
malinoff has quit
Diggsey joined the channel
mrud_ is now known as mrud
mrud has quit
mrud joined the channel
erve joined the channel
erve
I'm using 4.0.0rc2 (to take advantage of the much-improved SQS broker) and I've recently (in the last 48 hours or so) found that my workers are slowly dying after around 30-40 minutes of expected activity. I've been through all the issues/SO questions I can find. Tried all possible solutions. strace and lsof have shown me that hung workers are trying to read from a pipe to celery. Any ideas?
asksol
erve: hard to debug with the sqs broker since we don't have remote control commands
when you say hung workers are reading from a pipe, do you mean the worker main process, or the child processes?
the child processes are supposed to be reading from pipes when they don't have work