Monitoring

[Matt Molyneaux]

We need some URLs that Nagios and friends can connect to to check that Salmon and Celery are still running/working

For Celery, I think we can ping the workers as that happens over Rabbit (iirc). Alternatively, I could create a task that populates a database model ever hour. This might work better in situations where the workers are busy but still functioning.

I'm not sure what to do about Salmon. I could either check that an email has arrived in the last X hours or have a special address that records when an email has been received (but doesn't save the message). I don't like either of those solutions.

Directly checking the PID files and processes is out of the question as there's no guarantee they're running on the same server.