Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

We have a few DELL 1950 servers. One of those servers has CentOS6.3 and it's randomly rebooting, so I suspected it was hardware (no log generated). The other four servers do not randomly reboot.

We ran memtest86+ on the five servers and on three of them memtest86+ crashes (displaying an odd and colorful screen, like if a video card failed).

I tested an old memtest86 (not +) and none of the servers crashed. I also tested other RAM testing utilities, with no tool failing.

Have any of you guys experience this?

share|improve this question
    
Does it crash immediately? Or does the test run for awhile before it crashes? –  David Schwartz Dec 7 '12 at 19:22
    
it takes between 5-9 hours to reboot. (at least this last last week) –  user148723 Dec 7 '12 at 22:53
    
Then I would definitely suspect a hardware problem. If it crashed immediately, I'd suspect some compatibility problem with memtest86+. –  David Schwartz Dec 7 '12 at 23:06
add comment

3 Answers

If memtest crashes, there is a high chance that your memory is bad. Try to replace memory from non-crashing servers and re-running memtest. Most likely the memory is the culprit. You can also reduce the memory to half (system permitting; minimum memory requirements) and try running memtest. Once passes, try replacing memory with other half and see.

share|improve this answer
add comment

I you have a Linux server that is rebooting, this usually means that it is a hardware problem. Check the logs in Dell OMSA (Dell Open Manager System Administrator Managed Node). or via DRAC (Dell Remote Access Card).

Contact Dell technical support to assist you in the problem investigation.

share|improve this answer
    
thanks, I'm already talking to them about that. –  user148723 Dec 7 '12 at 22:52
add comment

Another tool to keep on hand that is exttremly useful to help test and diagnose is the UBCD. it include the memtest and memtest+. The new version even has a memory tester for GPUs. This will come in handy if you are suspecting the video having issues also.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.