Summary CPU Exception

From: <71055.111_at_compuserve.com>
Date: Mon, 08 Jun 1998 17:26:15 -0400

Thanks to all that responded, expecially to:

Kurt Carlson
soma_c_at_decus.fr (Claude SOMA - CNTS)
Arlene Smyth

Finally we replaced the memory module with problems and the problem
disapeared, no more cpu exceptions.

Below some of answers to my questions:

***************************************************************************
*******************
>Hello,
>
>Uerf is reporting me a CPU Exception, no panics, my computer is an
>AlphaServer
>2100, 512MB ( 4-128MB modules ), dual alpha 190Mhz, StorageWorks Raid
>Array
>200, with Raid EISA Host Adapter
>
>SRM console shows no error for any FRU ( show fru, show error...).
>
>Below are the output of uerf and dia.
>
>>From dia I obtained the following:
> "CPU Minor class 2. 660 Entry"
>
> it means a system detected error external to CPU. Probably ( surely )
> CPU is not the culprit, in my opinion and the following:

No, it's not... it's just the cpu that reports it (misleading messages
from dia|uerf).

>
>"Memory Module ID x00000000
>Error Register 1 x0004000100000000
> [Odd] Error Summary
> [Odd] EDC Corr Error"
>
>And error detected and corrected on Memory module 0, my first 128 memory
>module,
>so this look like the faulty fru.

It shows as a correctable error, likely a single bit error.
uerf lies about it being severe priority, it isn't.... Digital likely
isn't going to fix that lie in uerf since they're moving to dia.
These can be fairly common with 2100's, the correctable errors which
only become an issue if they start becoming frequent in which case
you'd want to schedule a preventative replacement.
As you've extracted above it is memory module 0.

>Below are the complete output of uerf and dia, any comment or idea
>will apreciated,

I'd suggest you just keep an eye on it for frequency and not worry about
low volumes. If you are concerned about the volumes and have a service
contract, contatc Digital.

Scripts to extract the information from uerf|dia in a human readable form
vs. the dia|uerf jumble can be found referencing ua_uerf with the kit:
        ftp://raven.alaska.edu/pub/sois/README.uaio
        ftp://raven.alaska.edu/pub/sois/man.ua_uerf
  kit: ftp://raven.alaska.edu/pub/sois/uaio-v2.1.tar.Z

kurt

_____________________________________________________________________
Kurt Carlson University of Alaska, ARSC snkac_at_java.sois.alaska.edu
(907)474-5763 910 Yukon Drive #108.84 Fairbanks, AK 99775-6200



***************************************************************************
******************

From: soma_c_at_decus.fr (Claude SOMA - CNTS)
 
I think this is a know bug at Digital on uerf
uerf claim CPUEXCEPTION but it's NOT, it' a memory error.

***************************************************************************
******************

Hello,

We had similar "CPU Exception errors", first of all we replaced the CPU
board
and then we replaced a bad memory simm and the latter seemed to fix the
problems

Hope this helps.

Arlene Smyth,
System Administrator,
Sheridan College.


***************************************************************************
******************

thanks again,

Abdon
Received on Mon Jun 08 1998 - 23:29:41 NZST

This archive was generated by hypermail 2.4.0 : Wed Nov 08 2023 - 11:53:37 NZDT