masters shouldn't be up and running for months at a time, they should be restarted periodically (that's another bug). We need nagios checks to assert that master processes are no more than X days old. I'm thinking 14 is a good number to start with. Leaving in releng until we figure out exactly what we want here.

(In reply to Chris AtLee [:catlee] from comment #0) > masters shouldn't be up and running for months at a time, they should be > restarted periodically (that's another bug). We need nagios checks to assert > that master processes are no more than X days old. I'm thinking 14 is a good > number to start with. An alert for any master with >14days uptime sounds good to me also. This will help us keep masters "fresh" by rebooting them. We can always adjust the 14 threshold up/down later based on how this feels after trying it. > Leaving in releng until we figure out exactly what we want here. I think thats all we need - anything else before we push over to IT for the nagios setup?

Do we still think masters shouldn't be running for so long?

doesn't seem to be an issue latetly