Dell Server & Error 31 on CPU0

[finchwizard]

I've got a Dell Server at work.

Dual 2.4GHz Xeons
2GB ECC RAM
3 x 32GB SCSI drives
Perc4 Controller.

I've got Debian installed on it at the moment, and occasionally, not often, I will get the following error:

Uhhuh. NMI received for unknown reason 31 on CPU0
Dazed and Confused, but trying to continue.

I'm not sure if this is a hardware problem or a Debian/Hardware problem.

Any help would be appreciated.

I searched around, couldn't find anything.

[drl]

Hi.

Brute force, googled "Dazed and Confused, but trying to continue. debian", found about 400 hits.

I have no idea what it means, but I like the message ... cheers, drl

[finchwizard]

Yea, I searched for that before, and went through quite a lot of the mailing lists with no real luck.

[drl]

Hi, finchwizard.

I do recall one report that had a satisfactory conclusion: they identified a bad stick of memory, replaced it, and the error never recurred. I don't recall which list it was in, regrettably ... cheers, drl

[drl]

Hi.

Here it is ... cheers, drl

...
Jan 13 16:43:53 picasso kernel: Dazed and confused, but trying to continue
Jan 13 16:43:53 picasso kernel: Do you have a strange power saving mode
enabled?
Jan 13 16:43:53 picasso kernel: Uhhuh. NMI received for unknown reason 21.

We checked the server's RAM, found a faulty module and replaced it.
Since then these messages vanished.
...
http://lists.us.dell.com/pipermail/l...ay/025840.html

[glaston]

When running KDE, I keep getting a strange(literally) error that starts shutting down apps.
It even spits it out through a remote ssh session.

The full error is::
Message from syslogd@wendigo at Mon Sep 29 09:15:24 2008...
wendigo kernel: Uhhuh. NMI received for unknown reason 31 on CPU 0.
AND
Message from syslogd@wendigo at Mon Sep 29 09:15:24 2008...
wendigo kernel: Do you have a strange power saving mode enabled?
AND
Message from syslogd@wendigo at Mon Sep 29 09:15:24 2008...
wendigo kernel: Dazed and confused, but trying to continue

ACPI is disabled in BIOS and not built into the kernel since this is a dedicated machine used for a file server and remote access gateway.

The only real result from a google search turns up this thread.

It IS an old machine.
Dual P111 800, 512MB RAM.
I built it in like 1998-99 as a design workstation for 3D modeling so it was actually a very capable machine in its day.
I've had slack running on it since 2002 and this never happened until I put slack12.1 on it.

I already replaced the memory modules with ones that I know are good.