lock disk failure

[Basso_1]

Helo guys.
at the syslog file i got this warning.
Nov 21 10:31:22 y cmcld: WARNING: Cluster lock on disk /dev/dsk/c4t0d0 is missing!

Nov 21 10:31:22 y cmcld: Until it is fixed, a single failure could

Nov 21 10:31:22 y cmcld: cause all nodes in the cluster to crash

i guess that it is a hardware failure of the disk lock.
i will need to replace the hard disk ( hard disk failure )
What is detailed the procedure ?
I have 2 lock disk in one volume group vglock with no data it is used only
for lock disk one of them need to be replaced.
THANKS FOR UR REPLY

[nanan]

Hi
just replace the lock disks and rewrite the vginformasion with vgcgfrestore.

Regards
nanan

[Stephen Doud]

The message does not mean there is a problem with the disk. The disk could be brand new- and Serviceguard would still report the problem if the cluster lock structure is missing from the disk.
The lock structure is applied to the disk using cmapplyconf. However, unless you re-create the cluster, cmapplyconf won't do it again (doesn't know the lock structure needs to be repaired).
Do one of the following:
1) halt the cluster, perform cmdeleteconf -f (caution - deletes the cluster), then activate the cluster lock VG and cmapplyconf all of the cluster and package configuration files.

2) open a software case with the HP Support Center and ask for the 'cminitlock' utility. This utility can be used while the cluster is running, to replace the cluster lock structure on the lock disk.

[Coolmar · Accepted solution]

Hi,

Your disk may not need to be replaced. There are other things to check first with this error. Please see the following thread for more ideas on troubleshooting before you replace the disk. In the following thread, it turned out the SAN people made an error that caused the problem. HTH,

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=793216&admit=-682735245+1165412204912+28353475

[Basso_1]

Dear Stephen i have open a case with HP support and we still waiting the solution the cluster is running normaly, it is a critical system so we cannot halt the cluster.
Thanks Coolmar for this link hope it will not be the same problem.
I will keep u updated with all the information.
REAGARDS