Moving from Lock disk to the quorum server

[jaivinder]

Hi Gurus,

I have a two node cluster. Lock disk was configured in this cluster.Now the lock disk is failed. But we do not want to change the lock disk, because we have configured a quorum server.Now i want the full steps to move the lock disk and to add the quorum server name in the cluster configuration files.

Thanks in advance.

[Frank de Vries · Accepted solution]

One handy document for disks:

http://docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad_WP.pdf

The cluster lock disk must be in an LVM VG that is seen by both nodes in the cluster (in /etc/lvmtab). As long as the VG is associated with the cluster, it is a legitimate cluster lock VG (and PV, since it owns LUNs). To change the cluster lock VG and PV designations in the cluster binary file, do the following:
1) cd /etc/cmcluster
2) cmgetconf CLS.ASC
3) edit CLS.ASC and locate:
FIRST_CLUSTER_LOCK_VG
and in each node section:
FIRST_CLUSTER_LOCK_PV
4) identify a new VG that will remain under cluster control after the SAN migration has completed
5) identify a PV in that VG (in /etc/lvmtab)
6) edit the CLS.ASC file, with the new lock VG reference and for each node, lock PV reference.
7) cmcheckconf -C CLS.ASC
8) If step 7 succeeds, halt the cluster and apply the configuration change
9) cmapplyconf -C CLS.ASC

[Duncan Edmonstone]

1. Get the current cluster configuration using cmgetconf:

cmgetconf > /etc/cmcluster/newclustercfg.ascii

2. edit this new file and add the following entries:

QS_HOST
QS_POLLING_INTERVAL 120000000
QS_TIMEOUT_EXTENSION 2000000

inserting either the hostname or IP address for the quorum server.

comment out the lines starting with the following text if they exist in the file:

FIRST_CLUSTER_LOCK_VG

FIRST_CLUSTER_LOCK_PV

SECOND_CLUSTER_LOCK_VG

SECOND_CLUSTER_LOCK_PV

save the file

Now halt all packages in the cluster:

cmhaltpkg

Once all packages are halted, halt the cluster:

cmhaltcl

Now apply the new configuration you have created:

cmapplyconf -C /etc/cmcluster/newclustercfg.ascii

Now restart the cluster:

cmruncl

That's it.

HTH

Duncan

[jaivinder]

Thanks for your replies.