Forums

Our account was put into a tarpit early this morning for going over our CPU allowance. When I logged into the account, I could see a whole bunch of running tasks jamming the system, including a large number of processes along the lines of:

/usr/bin/python -m virtualenvwrapper.hook_loader --script /tmp/virtualenvwrapper-initialize-hook-SnasTKxPKe initialize

I switched off all of our scheduled tasks, then killed all of the outstanding processes (scheduled tasks and the hook_loader tasks), and have begun to switch our scheduled tasks back on one-by-one. Unfortunately, they don't seem to be running, and the /usr/bin/python -m virtualenvwrapper.hook_loader tasks have begun to mount up again.

There's no obvious errors in the log files, apart from "Task failed to start. Too many processes/threads" on some of the scheduled task logs. The Django web-app itself is running smoothly, and I'm able to manually run the scheduled tasks via the command line. The only problem seems to be with the scheduled tasks and these .hook_loader calls using up our CPU...

Any ideas?

Sorry about that -- we did have some problems with one of our task servers this morning, and it was the one where your tasks run. Everything should be back to normal now...?

Thank you - yes, everything looks to be running normally again now.

Thanks for your help.

Excellent -- thanks for confirming!