booting issue (possible overheating?)

[jkay16]

I have a rack of servers and they are all on battery backup. Over the weekend we had a bunch of storms and the power went out. Our battery backups keep them online for roughly 2 or so hours. Our A/C does not have a battery backup, which usually isn't a problem except it was very hot and humid this weekend. I got a call that said our internet was down. When I got there 2 of the servers were on the bios screen with hundreds of ....................... and fans were at full speed. These two servers are the primary and secondary DC for our network. I switched all traffic to the 3rd DC so this isn't an problem. Anyways I turned them both off. Let them sit for a bit then turned the primary one on and it would boot into windows and about 10-20 seconds after it fully booted it would reboot and go back to the screen full of ......................

the secondary server booted and did the same thing except it stayed on for a few minutes. I was able to pull the drives out and shove them in other boxes and they booted fine and are working. So I know the drives aren't corrupted. Does anyone have any experience with this? Is this a overheating issue or did I burn something up inside the boxes? 

Thanks

[Suman_1978]

Hi,

What is the server model?

Any beeps?

If the servers are G7 and below, boot from SmartStart CD and perform diagnostics.

Have you connected any external storage devices which includes USB stick, then disconnect them and try.

Remove any unwanted PCI cards if installed (which is not necessary for booting).

Thank You!
I am a HP employee
__________________________________________
Was the post useful? Click on the white KUDOS! Star.

[jkay16]

It's a DL360 G4p....I think it was a overheating issue. I took the server out of the rack, cleaned it up and replaced the thermal paste and my issue went away. I think the cpu was getting too hot. I did not hear any beeps, it would just reboot after a few minutes of being on and get stuck in the bios screen with ..................................... for lines and lines of them and never do anything except fans were at 100% the entire time.

[waaronb]

The Proliant array controllers run really hot normally, so if your room lost cooling, I could see the SA 5i running REALLY hot (the G4 uses a 5i, doesn't it?  It's been a long time since working on those).

If the system itself had cooled down enough to where it would boot, the array controller might have still been overheated and just needed some nice power-off quiet time to chill out.

Replacing the thermal paste on the CPU was probably a good idea too since that stuff gets old and crusty with time anyway, and at high temperatures it ages faster.

The dots you saw, was that from the system trying to do a PXE boot?  That would tend to confirm it was the array controller that was overheated... the system booted, ACU was too hot and not working right, so the system tried doing a network boot?

Glad it's working now though... it's such a feeling of relief when a machine you thought was dead is back up and running.