Cluster Lock disk is missing

[Naj]

Hello expert,

I received error message "cluster lock disk is missing". 

May i know in why this happened and how to over come this?

Thanks

BR

Naj

[g3jza]

Hi,
what does #cmviewcl -v
say? In what state is the cluster lock disk?

Is the cluster lock disk's DSF seen in the ioscan output and is it CLAIMED / NO_HW?

Cluster lock disk is being used in SG clusters as an arbitrator device, to prevent split-brain situations in clusters....

[Naj]

Hi,

I can see there are under unknown state for 2nd node and current node is down state.

unfortunately when i try to wakeup the node, it wont and giving that error.

I perfomed ioscan and i can see all claimed but that disk mentioning on error message "disk is missing" not stated there.

i have down node A and working on node B to bring up the node.. is that good idea? or

Should i bring up both servers and run cmruncl ?

Thanks

BR

Naj

[g3jza]

Have you / somebody else been doing any changes in the SAN / storage ? Is the cluster lock disk on the same storage as the shared (application VG for packages) VGs? What is the storage?

What is the OS release, SG release?

So node A went down by itself?

Could you please post the output of /var/adm/syslog/syslog.log? Output of #ioscan -fn / ioscan -fnN

Output of
#cmviewcl -v

Any suspicious logs on the san switches?

[Naj]

Hi

Below is output

OS ver :11.3x

CLUSTER        STATUS
server002    unknown

  NODE           STATUS       STATE
  server002a   unknown      unknown
  server002b   down         unknown

UNOWNED_PACKAGES

    PACKAGE        STATUS           STATE            AUTO_RUN    NODE
    kd0010p      down             halted           enabled     unowned
    kd0011p      down             halted           enabled     unowned
    kd0012p      down             halted           enabled     unowned
    kd0007x      down             halted           enabled     unowned
    kd0008x      down             halted           enabled     unowned
    kd0009x      down             halted           enabled     unowned


Log message

Jul  4 22:30:34 server002b LVM[12750]: /usr/sbin/vgexport -s -p -m /etc/lvmconf/vglock.mapfile /dev/vglock
Jul  5 01:14:21 server002b sshd[18316]: SSH: Server;Ltype: Version;Remote: 138.35.215.45-44050;Protocol: 2.0;Client: OpenSSH_4.3
Jul  5 01:14:23 server002b sshd[18316]: Accepted publickey for ermclnt from 138.35.215.45 port 44050 ssh2
Jul  5 01:14:24 server002b sshd[18318]: SSH: Server;LType: Throughput;Remote: 138.35.215.45-44050;IN: 384;OUT: 78336;Duration: 0.4;tPut_in: 1044.9;tPut_out: 213155.1
Jul  5 01:30:35 server002b cimserver[1790]: PGS13606:  A failure was detected in provider module HPUXStorageProviderModule.  The generation of indications by providers in this module may be affected.  To ensure these providers are serving active subscriptions, disable and then re-enable this module using the cimprovider command.

Thanks

BR

Naj

[g3jza]

There's nothing useful in the syslog.log.... Only that someone exported the vglock VG in preview mode.

What actually happened?

Can you try to gather syslog.log from server002a?

Also post #cmviewcl -v from server002a .
You should start investigating with the help of your SAN / LAN team.

Some of my questions are still not answered....