Celery Jobs Profiling

Question

As I see, the toputility celeryuses a lot of processor time. Therefore, I want to view it.

I can do it manually on the developer's machine like this:

python -m cProfile -o test-`date +%Y-%m-%d-%T`.prof ./manage.py celeryd -B

But in order to have accurate timings, I need to profile it on a production machine. On this machine (Fedora 14), celery is started by init scripts. For instance.

service celeryd start

I found out what these scripts would eventually be called manage.py celeryd_multi. So my question is: how can I say celeryd_multilaunch celery with profiling? In my case, this means adding -m cProfile -o out.profparameters python.

Any help is greatly appreciated.

Answer 1

, . , .

, , .

, , celerymon. , , , .

, , , , . , , . .

cProfile.run('func()') , . run, , , , pstats.add, .

, , , , .