I had an odd Linux crash this morning that I just don’t understand. I got a call from Shannon saying that the webserver wasn’t responding; I tried to ping it, and it didn’t respond. In fact, no services (web, mail, FTP, ssh) were responding, so I walked over to the machine, and while it was running (I could hear the fan spinning, and the lights on the front were lit), I couldn’t get anything at all on the monitor. I ended up having to hit the reset button, after which it came back up just fine.

Looking through the logs, there is truly no clue as to what happened. The main log (/var/log/messages) shows that the IMAP server processed its databases successfully, and then the next entry is from almost six hours later, when I reset the machine. Every other log — mail, cron, security, webserver — shows the same thing: entries up until around 3:02 AM, and then silence for nearly six hours, until the machine is restarted.

This machine has been running fine for a while now — I’ve had eight-month stretches without reboots — but in the past three or four weeks, I’ve started installing all the services that will allow it to become my primary web and mail server. Maybe one of them took it down; without log information, it’s hard to know, though. I’ve started to do a dump of the memory state every five minutes or so… maybe that’ll give me some insight into what happened. Any other ideas?