A BIG KISS to all of you !!!!!!!!!
("Van Bever, Pascal", "Vallee, Mark", "Dr. Tom Blinn", "Sunil Kurupath", "Lavelle, Bryan",
"Kurt Carlson", "John Losey", "John Speakman", "Alan", "Nikola Milutinovic", "Werner Rost")
A special thanks to Sunil Kurupath from Compaq Corporation for his detailed
explanation.
The solution came in 10 minutes against 2 months of energy and time
consuming useless attempts.
SOLUTION :
==========
> This is a known problem and this starts after you install pk5 for 4.0D.
...... known, but not to everybody, e.g. our local Compaq support :-(
> Please install Patch Kit 6 or you can find a point patch to fix this problem
> at http://ftp.service.digital.com/public/Digital_UNIX/v4.0d/kzpba_v40d_bl13.tar
CONFIGURATIONS AFFECTED:
========================
> If ...
> you have multiple controllers (isp0, isp1, ... not just isp0, isp0, isp0)
> and you have multiple cpus
> (grep -i 'secondary cpu' /usr/var/adm/messages)
> and you are running V40D/E/F BL13,
> then the system is a definite candidate for this blitz.
>
> This problem will occur on system configurations as described above
> with the BL13 V4.0* patch releases installed:
>
> 4.0D DUV40DAS0005-19991007 Patch Kit 5
> 4.0E DUV40EAS0003-19991110 Patch Kit 3
> 4.0F DUV40FAS0002-19991116 Patch Kit 2
My original question was:
> Hello,
>
> we are getting mad with an apparently unsolvable problem manifested by
> CAM SCSI errors in the binary.errlog file in correspondence with dump/vdump
> backup of the disks on tapes (DAT).
>
> We did not have such problems before the end of last year, using the same
> hardware. However there was the y2k problem, you know :-) , and we upgraded
> from 4.0B to 4.0D + patches + firmware upgrade. After that we could make just
> a full dump without problems and then, the sequence of errors started.
>
> We used different dat devices (and different cables).
> Compaq support, after checking the SCSI HW (KZPBACX), suggested that the
> problem could come from some possible corruption of the system SW. After
> verifying that if we boot the system from CD there is no problem in dumping
> fylesystems on tape, we have gone through a painful fresh install of 4.0D +
> patches and now... we get again the same errors.
>
> At this point we look for any suggestion/help in understanding what could be the
> origin of the problem. Is there any way to understand it from the uerf output ?
> I enclose some information about the present HW/SW configuration as well as a
> sample output from uerf -R -o full.
>
Alessandra Richetti
_________________________________________________________________
-- Alessandra Responsabile Elaborazione Dati
/__\ Richetti Dipartimento di Fisica Teorica
/(__)\ Universita` degli Studi di Trieste
(__)(__) Strada Costiera 11 - 34100 TS - Italy
-----------------------------------------
e-mail: richetti_at_ts.infn.it
url:
http://www-dft.ts.infn.it/
phone: +39 040 2240299 fax: +39 040 224601
Received on Wed Feb 23 2000 - 08:14:18 NZDT