change cluster lock disk

[RiclyLeRoy]

I need to change cluster lock disk in service guard which manges 3 hosts because I have to change lock disk.

What steps I need to follow to make it?

I need to activate lock volume group and and to its replace physical volume  ? Or I need to change service guard cofiguration ? Or both ?

Thanks

[Sush_S]

Hi RiclyLeRoy
Would you please let me know the Serviceguard version that you are currently using?
Thanks,
Sush_S

[RiclyLeRoy]

@Sush_S 

ServiceGuard version is A.11.19.00

[Sush_S]

Hi RiclyLeRoy,
Access the following video link for a detailed demonstration of the steps needed to carry out this task.
https://support.hpe.com/hpesc/public/videoDisplay?videoId=vtc00000875en_us 
(For clusters with more than two nodes, it is recommended to utilize a quorum server)

Thanks,
Sush_S

[RiclyLeRoy]

I saw video and I tried to run this command:

cmgetconf cluster.config

I got error the following error "Unable to receive device query message from hostXX: Software caused connection abort" for every nodes.

This issue is shown on all nodes.

[georgek_1 · Accepted solution]

Hello  RiclyLeRoy,

The message you are getting could be due to inconsistency with volume groups configued in SG nodes .
Check syslog.log and see if there are any referance to any of the VGs , make sure that it is properly configured in all the nodes .
Please check the vgs configured (SG vgs) in each node and make sure that all nodes are updated with correct devices in all the nodes .

As a workaround, you may use cmgetconf with -K while running cmgetconf/cmcheckconf and cmapplyconf .

Note : Make sure that you have mentioned the Cluster aware VGs in output file - manually,  before running cmcheckconf/applyconf .
           -K        This option causes the probing of volume groups to be
                     skipped, reducing the time for the command to complete.
                     Note that if this option is used, cmgetconf will not
                     add the list of cluster-aware volume groups to the
                     resulting cluster configuration file.  You must add the
                     cluster-aware volume groups to the cluster
                     configuration file before running cmapplyconf, or use
                     the -K option of cmapplyconf when applying the
                     generated configuration file.  Failure to do so will
                     result in all volume groups in the cluster being
                     unusable in packages or for the cluster lock.

I work for HPE/ I am an HPE Employee (HPE Community)