The issue had to do with PAM. I was getting zombie processes.
adding 'args = blocking=yes' to the passdb pam area fixed the issue.
Thanks for the response.
Patrick
Kyle Wheeler wrote:
On Sunday, August 19 at 02:13 PM, quoth Patrick - South Valley Internet:
Now that we're in the production environment, we've noticed that every 20 minutes, Dovecot will stop running.
Meaning what? Is the dovecot process still alive? Is the service unresponsive? Is it just not allowing logins?
Is the shared NFS store the only thing shared between the two? Is there a load-balancer in front of the machines?
I can't see anything out of the ordinary in the log files either.
What are the last couple entries in the log files when it "stops"? The same things every time, or something different every time?
~Kyle