I/O fencing fails

[ytv]

Hi,

Scenario is as following:

1. We have 2 3-node clusters with indepedent storages.

2. We are trying to remove a node from cluster2 and add the same node to cluster1.

3. To configure fencing, have edited /etc/llthosts on all nodes, /etc/llttab on the new node, /etc/gabtab on nodes.

4. Started fencing using command /etc/init.d/vxfen start

5. And when I try to start / reconfigure / check fencing, getting an error.  Following is the extract for reference.

bash-3.2# vxfenconfig -c

VXFEN vxfenconfig NOTICE Driver will use SCSI-3 compliant disks.

VXFEN vxfenconfig ERROR V-11-2-1006 List of coordinator disks in running

        cluster is different than local node.

        Unable to configure vxfen.

Following messages were noticed in /var/log/messages.

Nov  5 23:35:15 blrhpcmd05 kernel: GAB INFO V-15-1-20036 Port b gen   f7620a membership 0 2

Nov  5 23:35:15 blrhpcmd05 kernel: VXFEN ERROR V-11-1-23 Got the serial number of 600601609AD028004679F848C6D3E011 for coordinator disk

Nov  5 23:35:15 blrhpcmd05 kernel:      from remote node. Corresponding value for local node is 6006016031D228005A0FFEDCDED6E011.

Nov  5 23:35:15 blrhpcmd05 kernel: VXFEN ERROR V-11-1-11 Snapshot different. Dropping out of cluster.VXFEN INFO V-11-1-35 Fencing driver going into RUNNING state

Nov  5 23:35:15 blrhpcmd05 kernel: GAB INFO V-15-1-20032 Port b closed

Please provide me a procedure to resolve this online without affecting the current cluster functionality.

Thanks in advance.

[Tmy_70]

It is important to verify that the disks and disk array support SCSI3 PR. Read the man pages on vxfentsthdw and vxfenadm.

Please, check the link:

http://www.symantec.com/business/support/index?page=content&id=TECH77671

[Yasuhisa_Ishika]

It seems that corrdinator disks presented to this node is different from those of other nodes. Serial number
Have you changed disk configuration so as to let the node have a access to coordinator disks of new cluster when it joins to new cluster?

[AHerr]

You could also verify that the SCSI3 disks support updating keys by running the vxfentsthdw command.  It validates your environment and ensures that the disks are supported.  This test will erase any data on the drive so only do this on your coordinator disks or empty disks.