FreeBSD 6.3 Server Crashing - full /tmp?

[cerulean]

Is it possible to bring down a FreeBSD box by having a full /tmp folder? I've had frequent crashes the past week on one of my servers (which up until then had uptime in the 200+ day range). I thought it was hardware related (usually is) but after running some diagnostics (hard disk fitness test, memtest86, cpu burn-in, monitor incoming power levels, ambient temperature/internal cooling, etc) everything looks ok and I can't seem to locate an issue.

Unfortunately nothing is showing up in the logs regarding any type of kernel panic so it appears the server was just restarting. I *think* a full /tmp would not cause this (at most, core dumping applications which has occurred) but figured I'd ask. At this time, I'm thinking the only option I have is migrating to completely new hardware but figured it wouldn't hurt to ask the community to see if I might be overlooking something. Thanks!

Joe

[anomie]

I don't have a FBSD 6.3 box sitting around at the moment, but I would be interested to see if you (or someone) could recreate a crash by filling up /tmp. Like you, I'd be surprised if it brought everything down.

[cerulean]

After doing further diagnostics on the hardware, I'm pretty confident it is a thermal related hardware issue. I'll be adjusting the airflow in the box to see if that resolves the issue (at this time, I'm not quite sure exactly which component is the source of the errors). Doesn't seem to be a full /tmp issue.