Hi, I've got a bunch of servers running ISP Config (they vary between 3.0.4 and 3.0.5 from SVN) and they all seem to eventually run into this issue. The issue is that at some point email will start to stack up in the mail queue, and is no longer sent. The reason is because amavis is no longer running. I'm not able to locate any error logs that have any information relating to amavis, and the only indicator that something has gone wrong (aside from emails not sending) is that the mail logs are reporting connection refused to the amavis port. I'm not sure what the common denominator is because it's a very gradual thing. I had one server running for 150 or so days until the issue cropped up, and another that ran for only a day or so. All of the servers are a standard ISP Config install on Debian 6. Any help would be appreciated.
How much ram does the servers have? I run the setup used by ispconfig for years now and never had a problem with amavis yet. Amavis is also not restarted or reconfigured by ispconfig when mail users or domains get added, so it can not hang on a reload. To find out the reason you can try to install munin and monit to monitor your system parameters and to restart amavis in case it hangs.
The servers vary between 512MB to 4GB of RAM, all of them exhibiting the issue eventually. I believe I have Munin running on a couple of the servers, so I'll see if I can track down some info, if not I'll get it up and running on a couple of my servers. Unrelated: Do you prefer to receive these kind of troubleshooting requests here or at the issue tracker in your signature? Or are they for different things?
