This is an idea to try - I will do it.
In a couple of weeks I will tell back if it could help to
localise the problem.
You seem to have experiences with CRCD:
I assume CRC errors will both be reported in ACTLOG
and cause I/O in trouble to be automatically repeated,
is both true?
best regards
Juraj
-----Ursprüngliche Nachricht-----
Von: Zlatko Krastev [mailto:acit AT ATTGLOBAL DOT NET]
Gesendet: Mittwoch, 06. August 2003 12:16
An: ADSM-L AT VM.MARIST DOT EDU
Betreff: Re: virtual write error on Raid-1
Juraj,
have you tried `upd stg RECLAIMTPDB crcd=y`? It also might be worth to
enable it for few days on tape pool as well.
Note: it might slow down your operations and LTO is sensitive to this!
Zlatko Krastev
IT Consultant
Salak Juraj <j.salak AT ASAMER DOT AT>
Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
06.08.2003 12:29
Please respond to "ADSM: Dist Stor Manager"
To: ADSM-L AT VM.MARIST DOT EDU
cc:
Subject: virtual write error on Raid-1
Hello all,
anyone has seen before?
TSM Server reports sometiems write error on a disk raid-1 volume,
while
neither an evidence about an error is elsewhere (Raid-Controller, event
log)
nor I am able to provoke an write error wit a test program.
This problem occures sometimes, but not always,
when an reclamation process reclaimss an LTO tape to a disk volume
defined in
2003.08.05 11:41:50 ANR1340I Scratch volume
V:\TSM\RECLAIMPOOL1\00000DAA.BFS
is now defined in storage pool RECLAIMTPDB.
2003.08.05 11:47:11 ANR9999D pvrfil64.c(1056): ThreadId<30> Error
writing FILE
volume V:\TSM\RECLAIMPOOL1\00000DAA.BFS.
2003.08.05 11:47:11 ANR1411W Access mode for volume
V:\TSM\RECLAIMPOOL1\00000DAA.BFS now set to
"read-only"
due to write error.
TSM 5.1.7.0 on W2K SP3
IBM 345 with two Raid 4m controller, RAID-1 Disks
|