where to look to troubleshoot crashes

[sde]

i have a debian box running 2 half-life servers, ftp, apache, and everything else is pretty much standard. it's a dual amd system with a 120gb hd.

i'm experiencing crashes. i don't know where to look to track down this problem.

since the server is co-located, so i have to call to have them restart it.

it was up for about 8 hours yesterday before the crash.

any ideas on where i can look?

[bdl]

I wouldn't think you'd need to restart the machine, unless you want to compile and run a new kernel. How long has this specific config been running? First day, only 8 hours? Or has this machine been up and running for awhile? It's a custom machine that has been tested and 'burned in' prior to installation at the host?

Best bet, since you have to rely on doing things remotely, is start taking a look at whats running and how much memory you're taking up. What is the most intensive app on the system? What percentage of all applications are being utilized? What do the logs say around the time of the crash? Anything specifically happening at that time?

Barring software / system software problems, you'll need to take a look at the hardware; CPU(s) may be running too hot due to lack of cooling, memory is not seated properly, etc.

[sde]

thanks b, .. now that you mention it, perhaps it is a cooling problem. i didn't think of that... but that would explain it.

half life is the biggest resource user, up to 25% cpu and 13% memory. i'll keep an eye on that process.

i don't know what i was thinking when i posted in the lounge .. haha.