I bought a replacement supermicro motherboard off fleabay that has dual
Opteron 250 @2.4GHz. The cpus have passive heatsinks, it is in a supermicro
2U chassis with 4 front fans. After a couple of minutes of running at
essentially idle (I was in single user mode reconfiguring /etc/fstab to
compensate for my LSI MegaRAID card being found *after* the external enclosure
instead of before) when I began to hear an alarm and then shortly later the
server shut down. The rear cpu heatsink was noticeably warmer than the front one.
I let it cool and then when I restarted it I went directly to the system
health tab in the bios. The first cpu was running at about 50 deg C and the
second was in the hight 60s and climbing. The thermal cutoff point was set at
71. The amd specs (if I've found the right model in the pdf) say that the
maximum is 70 or 71 deg C. I shut the server down before it got to the
thermal cutoff point.
Opterons are new to me. Have I already damaged the CPU? I can get an couple
of active CPU heatsinks to replace the passive ones but if that chip is
already damaged I'd rather lose some more time and return the motherboard
while I still can.
If I do put on the active heatsinks will that generally be enough to bring the
temp back down to normal operating range? I thought I'd pull the passive
heatsink tonight and clean and then reapply the thermal paste to see if that
makes a difference.
Thanks,
Jeff Ross
- Opteron 250 Overheating Jeff Ross
-
