Topic: Server periodically unresponsive

[Marc Pfister]

Our SME file server, which has been flawless for over a year, is now becoming unresponsive once or twice  a day.

/var/log/messages gives no clues. There's nothing logged that would indicate any problems. Dmesg is normal too.

I suspect the power supply is the problem, but are there any other logs or things I should check on the software side to troubleshoot this behavior?

Thanks,

Marc

[Reinhold]

Marc

Have a look at the sensor capabilites of your mainboard... (Manual/Bios Settings)
If the mainboard is able to measure voltages then install the sysmon contrib including lm_sensors. (Shad Lords, Ian Wells).
You get a nice graphical view that would reveal most PS and Fan problems (+ a lot more).

BTW: Check your RAM! .-) - Good luck

[Marc Pfister]

I tried to install lm_sensors in order to get temperature readings, but when I modprobe the driver it doesn't find any sensors. I haven't tried to get voltage readings - I'll look into that.
I did notice one of the fans on the back of the case is not working, but it only draws air over the raid array. I disconnected it in case it had burned out and was having intermittent shorts.
Yesterday the system was crashing as often as every hour. It shows now log messages, kernel panics, etc.
Right now I'm running the machine off of a Knoppix CD just to completely rule out software issues. A new power supply is coming in today too.
I'll try a RAM test today too.

Thanks,

Marc