Environment
Novell Open Enterprise Server 2 (OES 2) Linux Support Pack 3
Situation
Cluster resource hangs and goes comatose while executing command
"nss /pooldeact"
The resource unload.out shows the last command as "nss /pooldeact="
The resource unload.out shows the last command as "nss /pooldeact="
Resolution
A fix has been made in the novell-ncpenc rpm.
Released as part of July 2012 OES Maintenance Update patch (oes2sp3-July-2012-Scheduled-Maintenance)
Released as part of July 2012 OES Maintenance Update patch (oes2sp3-July-2012-Scheduled-Maintenance)
Cause
There is a race condition between NCP (which runs in ndsd process
space) and NSS where if there is outstanding trustee
synchronization work that needs to be done when the NSS pool
deactivation process begins, there is a potential to see this race
condition. Restarting ndsd at this point does alleviate the
condition and allows the pool deactivation to
complete.
Additional Information
Inconsistent NCP trustee database for a volume contributes to this
problem. Please make sure the trustee database is in sync
with the NSS trustee database. You can run the
following command to check consistency.
ncpcon nss verify=<volumename>
Depending on the results, you can run the resync command to get these back in sync.
ncpcon nss resync=<volumname>
ncpcon nss verify=<volumename>
Depending on the results, you can run the resync command to get these back in sync.
ncpcon nss resync=<volumname>