Our vBulletin migration is complete.
Welcome vBulletin users! All content and user information from the Micro Focus Forums (vBulletin) site has been migrated to this site. READ MORE.
afirizarry_uw Absent Member.
Absent Member.
711 views

CIFS enabled cluster resource mounts on one of two nodes

OES2015sp1. Two node cluster. CIFS enabled cluster resources. Have had no problems migrating resources back and forth.
Migrated resources from node2 to node1. Rebooted node2 and it rejoined the cluster. Migrated a resource back to node2 and the resource would go comatose. See this in the cifs.log:
Dec 19 20:35:16 NODE2 CIFS[10899]: EVENT: CLI: Removing virtual server ".cn=XXXXX-SERVER.ou=XXX.o=XXXXXX.t=XXXXXX_TREE." with IP: 0x55d06880
Dec 19 20:35:16 DATA2 CIFS[10899]: WARNING: CLI: RemoveVServer: Failed to find virtual server "XXXXX-SERVER.XXX.XXXXXX.XXXXXX_TREE", NetBIOS name: XXXXXX-W in local list, probably already got deleted

Have no problems loading the resource on node1, but will go comatose on node2
Labels (2)
0 Likes
3 Replies
Knowledge Partner
Knowledge Partner

Re: CIFS enabled cluster resource mounts on one of two nodes

Check load and unload logs in /var/opt/novell/log/ncs/ on BOTH nodes. The issue could really be with the node where the resource works properly. On unload most commands are set to ignore errors (i.e. process proceeds even if something has failed) while load scripts are for obvious reasons more picky. So if something hasn't been cleaned up properly on unload it might cause problems on load on an otherwise "not guilty" node.
0 Likes
Knowledge Partner
Knowledge Partner

Re: CIFS enabled cluster resource mounts on one of two nodes

On 20.12.2018 04:14, afirizarry uw wrote:
>
> OES2015sp1. Two node cluster. CIFS enabled cluster resources. Have
> had no problems migrating resources back and forth.
> Migrated resources from node2 to node1. Rebooted node2 and it rejoined
> the cluster. Migrated a resource back to node2 and the resource would
> go comatose. See this in the cifs.log:
> Dec 19 20:35:16 NODE2 CIFS[10899]: EVENT: CLI: Removing virtual server
> ".cn=XXXXX-SERVER.ou=XXX.o=XXXXXX.t=XXXXXX_TREE." with IP: 0x55d06880
> Dec 19 20:35:16 DATA2 CIFS[10899]: WARNING: CLI: RemoveVServer: Failed
> to find virtual server "XXXXX-SERVER.XXX.XXXXXX.XXXXXX_TREE", NetBIOS
> name: XXXXXX-W in local list, probably already got deleted



That doesn't really fit together. That error you're showing happens when
it tries to *unload* on NODE2. Either there is an earlier error in the
loading of the ressource which causes it to try and unload it, or that
error in some other way isn't related to a problem *loading* the
ressource on Node2.

Maybe post your load and unload scripts?

CU,
--
Massimo Rosen
Micro Focus Knowledge Partner
No emails please!
http://www.cfc-it.de
0 Likes
afirizarry_uw Absent Member.
Absent Member.

Re: CIFS enabled cluster resource mounts on one of two nodes

Problem has been resolved. It is a bit more complicated than I explained but following the resource load logs did lead me to the solution early this morning.
The cluster resource having problems has one NSS pool and volume. We recently set the volume up for DST so a second LUN was created and a new pool/volume created for the secondary or shadow volume for DST. That second LUN was on an iSCSI disk. Looking at the logs I finally realized that the problem was when the cluster load script attempted to activate the pool used by the shadow volume. It was failing and going comatose at that point. I compared the output of nss /pools on the two nodes and noticed that the pool for the shadow volume did not appear on the impacted node. I compared the devices in nssmu and noticed that the impacted node did not see the device associated with the shadow pool/volume. I then checked the iSCSI initiator in YAST and found that the service was not set to auto start upon boot. I set it to auto-start, rebooted the node and after coming back up I could see the device and pool. The resource then migrated just over just fine.
I was fixated on CIFS because we have had issues in the past with CIFS errors causing resources to go comatose when migrating resources between nodes. The issue has gone away with recent OES patches but it caused me to believe the CIFS problems were back and waste a lot of time barking up the wrong tree. Thanks for the replies.
0 Likes
The opinions expressed above are the personal opinions of the authors, not of Micro Focus. By using this site, you accept the Terms of Use and Rules of Participation. Certain versions of content ("Material") accessible here may contain branding from Hewlett-Packard Company (now HP Inc.) and Hewlett Packard Enterprise Company. As of September 1, 2017, the Material is now offered by Micro Focus, a separately owned and operated company. Any reference to the HP and Hewlett Packard Enterprise/HPE marks is historical in nature, and the HP and Hewlett Packard Enterprise/HPE marks are the property of their respective owners.