17.7.3 Unconfiguring NNMi on the active cluster node

On the active cluster node, stop the NNMi HA resource group:
- Windows
%NnmInstallDir%misc\nnm\ha\nnmhastoprg.ovpl \

NNM resource-group
- UNIX
$NnmInstallDir/misc/nnm/ha/nnmhastoprg.ovpl \

NNM resource-group

Because this command returns the prompt without waiting for the resource group to stop, use an HA cluster software or management console command to verify that the resource group has stopped.
Depending on the environment in which NNMi is used, you might need to perform the following procedure:
- WSFC
  
  After you have stopped the resource group, follow these steps from the Failover Cluster Management console:
  
  Delete the disk resource from resource-group-APP dependencies.
  
  Remove the disk resource from resource-group.
On the active cluster node, unconfigure NNMi from the HA cluster:
- Windows
%NnmInstallDir%misc\nnm\ha\nnmhaunconfigure.ovpl \

NNM resource-group
- UNIX
$NnmInstallDir/misc/nnm/ha/nnmhaunconfigure.ovpl \

NNM resource-group

This command removes the corresponding node from the list of failover targets in the HA resource group.

This command removes access to the shared disk but does not unconfigure the disk group or the volume group.

The following message might be displayed, but this is not a problem:
- WSFC
  
  Warning: There is no value in the public entry PUBLIC.HA_MOUNT_POINT in the resource group xxxxx in the cluster registry.
- HP SG
  
  Unable to perform the security token exchange with cmclconfd on node xxxxx
  
  Cannot connect to configuration daemon (cmclconfd) on node xxxxx
- VCS or SCS
  
  VCS WARNING V-16-1-10133 Group does not exist: resource-group
Change the FQDN setting for the NNMi node to the physical host name.

For FQDN, specify the FQDN of the physical host name (host name displayed by the hostname command).
- Windows
  
  %NnmInstallDir%bin\nnmsetofficialfqdn.ovpl -force FQDN
- UNIX
  
  /opt/OV/bin/nnmsetofficialfqdn.ovpl -force FQDN
This command changes the FQDN setting that was changed to the virtual host name during HA configuration to the FQDN of the physical host name.

The following messages might be displayed during command execution:
- For single sign-on to function normally, you must generate a new certificate manfully
  
  Ignore this message because single sign-on is not supported.
- A new certificate cannot be generated. The self-signed alias xxx.xxx.xxx already exists in the key store
  
  Edit the nms-local.properties file.
File path

Windows: %NnmDataDir%conf\nnm\props\nms-local.properties

UNIX: /var/opt/OV/conf/nnm/props/nms-local.properties

What to edit
- If you use self-signed certificates
  
  com.hp.ov.nms.ssl.KEY_ALIAS = FQDN.selfsigned
  
  For FQDN, specify (in lowercase) the FQDN specified in nnmsetofficialfqdn.ovpl.
For details about the certificates, see Chapter 8. Working with Certificates for NNMi.
On the active cluster node, move the files that are specific to the NNMi HA resource group to a separate location for safekeeping.

If you do not plan to reconfigure the NNMi HA resource group, you do not need to save a copy of these files, and you can delete them at this time.
- WSFC
  
  In Explorer, delete the %NnmDataDir%hacluster\resource-group\ folder.
- HP Serviceguard
  
  cd /var/opt/OV/hacluster
  
  rm -r resource-group
  
  cd /etc/cmcluster
  
  rm -r resource-group
- VCS or SCS
  
  cd /var/opt/OV/hacluster
  
  rm -r resource-group
Mount the shared disk.

Use an OS or cluster operation to make the shared disk accessible.
Copy the NNMi files from the shared disk to the previous active cluster node.

Perform this procedure if either of the following is applicable:
- You will be running NNMi with a database migrated from an HA configuration to a single-server configuration.
- You used nnmchangeembdbpw.ovpl in the HA configuration to change the database password.
Use the following command to copy the NNMi files from the shared disk to the local disk on the previous active cluster node:
- Windows
  
  %NnmInstallDir%misc\nnm\ha\nnmhadisk.ovpl NNM -from HA-mount-point
- UNIX
  
  /opt/OV/misc/nnm/ha/nnmhadisk.ovpl NNM -from HA-mount-point
Delete the NNM folder or the NNM directory from the shared disk.
Unmount the shared disk.

If you will be running NNMi on the previous active cluster node, the preparations are now complete.

Execute ovstart to start NNMi.

The information provided below applies if either of the following is applicable:
- You will be running NNMi on the previous passive cluster node using a database migrated from an HA configuration to a single-server configuration.
- You used nnmchangeembdbpw.ovpl in the HA configuration to change the database password.
Use the command shown below to back up the NNMi configuration on the previous active cluster node.

This provides a backup containing the data that was copied from the shared disk to a local disk in step 7.
- Windows
  
  %NnmInstallDir%bin\nnmbackup.ovpl -type offline -scope all -target directory
- UNIX
  
  /opt/OV/bin/nnmbackup.ovpl -type offline -scope all -target directory
Restore the active cluster node backup data obtained in step 10 to the previous passive cluster node on which you want to run NNMi using the data that was used with the HA configuration. Do this on one node at a time.
- Windows
  
  %NnmInstallDir%bin\nnmrestore.ovpl -force -source backup_data
- UNIX
  
  /opt/OV/bin/nnmrestore.ovpl -force -source backup_data
For details about this command, see Chapter 18. NNMi Backup and Restore Tools.

To Page Top