ProLiant 2500R/RH linux 7.2 traumatic power off

[Angus Morrison]

A Proliant 2500R (standing on edge as per the tower model) single PP 200 processor runs for about 3 hours then stops witout warning. It will attempt to power on after about 5 seconds, run for one second then stop. After switching off for about 12 hours it will start again.

The contacts on the memory (2 x Legend 256 Mb) have been cleaned and chips reseated. contacts on the processor and PCI daughter cards have been cleaned and re-seated.

Any help would be appreciatted.

[JohnWRuffo · Accepted solution]

Angus:

What you are experiencing likely is a Thermal Shutdown by the ASM (Automatic System Recovery). The next time you have the system running, make sure all the fans are turning and that you do not hear any "excessive" noise from them. Also, make sure there is plenty of room in front and behind the server to allow for good airflow. Your system may just be getting too hot and shutting down.

When you "cleaned" the inside of the box, was the cleaning to remove dust buildup? If so, you can shutdown the systems and unplug it and "blow out" with a can of air the power supply and the front covers of any residual dust. Make sure to get it all out with a vacuum.

[Angus Morrison]

John,

Thanks.

All dust has been removed; the panel on the front cover hiding the fan for the PCI section has also been removed and the server has been running for over 4 hours so far without problems. The fans at the rear are blowing cool air. The problem appears to have been overcome.

[SAKET_5]

Hi there,

A couple of points:
1. Is your system BIOS up-to-date?
2. Are the OS drivers up-to-date?
3. Could you check the hardware log (Integrated Managament Log IML) and find out if the system is experiencing any ASR (Automated System Recovery) events.
4. Could you also check if the temperature of the system is under the tolerable limit? You can do this by either checking the system temperature at the BIOS level or through management agents if you are running them.
5. Could you also check the temperature of the room where the server is placed is under control. No doubt it would be A/Ced but we have noticed that even in powerful ACed rooms, sometimes there are hot spots. Is the server getting affected by any of that?
5. Could you also check if there are any empty PCI slots without covers? As mentioned before I would also check if there is not too much dust in the internals of the server or around it.

Interested to hear your comments.

Good luck.

[Prashant (I am Back)]

Hi,

I am not good at linux but what i can suggest is we can install psp for linux so we can check for any previous error logs on server why it went down.

Or

If require i can provide offline IML DOS reader to get the same details.

Regards,
Prashant S.