NSS Error after upgrade - Visibility errors - POOLREBUILD?

Question

(Question at bottom)
I upgraded a remote NW6.5sp4a server on the weekend (was having no
problems with it but did not realize it was on such an old service pack)
Server was 6.5SP4a - 5.70.04
DS - 10552.79
NSS - 3.23-2
All NSS volumes, no traditional anywhere
Pools: SYS, DATA
SYS Pool Volume: SYS 8GB Total Size - 5GB Free 36MB purgable
DATA Pool Volumes: DATA, CSG, USERS, ADMIN
The server had transactions enabled on SYS which I disabled before the
upgrade.
I did a unattended DSREPAIR which was clean
Installed SP8, iprntnw65sp8a,and TCPIP--NW65SP8 then rebooted server.
Server came back up and was basically fully booted and then went into
the debugger.
I downed the server from the debugger and restarted at which time it did
complete the load.
From here my PRODUCTS.DAT file was hosed but I copied in one from
another SP 8 server and rebooted again. Everything seemed to come up
fine and I could see items in the PRODUCTS.DAT again.
I then installed edir87310_ftf2 and rebooted
then installed N65NSS8b and rebooted.
At this time I did another unattended DSREPAIR which was clean
I did a full purge on all volumes from a workstation.
Things were fine but when I checked it on Monday morning the logger
screen was showing the following error (BackupExec running backup when
it was logged)
1:39:04 am NSS<COMN>-3.27b-714: zasAuthModel.c[2057]
Error removing visibility for trustees just removed
I was then of the thought that I would need to do a
nss /visibilityrebuild this coming weekend to fix this error.
In the meantime I decided to run CHECKVIS.NLM /SYS to see what it would
return. I was led to believe that this would have minimal impact to
users but it rendered the server unavailable. I could ping it and it
would respond and I could move between console screens but that was it.
We ended up trying to down the server through the alternate console
which did not work. I then downed it through the debugger and restarted it.
When loading it came back with the following errors on the console
Loading volume "SYS"
Activating volume "SYS"...
** Volume layout v36.03
** Volume creation layout v36.03
** Processing volume purge log
** ......Error fixing visibility of trustees that were being added or
removed when the system crashed
Still have the error showing up when the backup runs
Mar 11, 2010 1:39:04 am NSS<COMN>-3.27b-714: zasAuthModel.c[2057]
Error removing visibility for trustees just removed
Questions:
Is it normal for the CHECKVIS to render the server unavailable?
How long should the CHECKVIS take on and 8GB volume (15K 320SCSI Mirror)
How should I fix this? Should I just run a POOLREBUILD /PURGE on SYS
and then the other pools? NSS /VISIBILITYREBUILD as well?
Thank you for any assistance you can offer!
Regards,
Dieter

Answer 1

Dieter wrote:
>How should I fix this? Should I just run a POOLREBUILD /PURGE on SYS and
>then the other pools? NSS /VISIBILITYREBUILD as well?
Yes for both. Make sure you have a backup of your data just in case.
Note that the time a rebuild and a visibility rebuild doesn't depend so
much on the volume size but rather on the number of objects (files &
directories) that reside on your volume. However 8GB is pretty small and a
rebuild should only take a few minutes in this case.
Marcel Cox
http://support.novell.com/forums
Marcel Cox's Profile: http://forums.novell.com/member.php?userid=8