Right now, if we can't talk to RabbitMQ, Celery will (using its defaults) keep trying for 4 seconds * 100 retries, or 400 seconds. That is entirely too long. I'm inclined to say a 2-second timeout is more than enough, and zero retires. The whole point of Celery is to not take a long time on the web-app thread. We should have some other monitoring, like nagios, for RabbitMQ connectivity. If it dies, it shouldn't prevent access to the site.
We want to add more monitoring before making this change.
Target Milestone: 2011-10-11 → 2011Q4
Cleaning up 2011Q4
Target Milestone: 2011Q4 → ---
At this time we don't think there is enough information to move forward with this bug. Either this isn't an issue or it's been fixed. I'm going to mark this as INCOMPLETE.
Status: NEW → RESOLVED
Last Resolved: 5 months ago
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.