Description

When migrating Coordination Point Servers (CPS) to new servers with a new operating system, it is important to ensure that the CPS UUID file is copied over correctly. If the CPS UUID file is not copied, it can cause issues with fencing and cluster node reboots. This guide provides step-by-step instructions on how to replace the CPS UUID file to resolve these issues.

How To Replace CPS UUID File for Coordination Point Servers Migration:

1. Verify the current CPS UUID file:

   • Log in to the new CPS server.
   • Open a terminal or command prompt.
   • Run the following command to check the current CPS UUID file:
     
     ll /etc/VRTScps/db/current
      
   • Note the name of the CPS UUID file (e.g., cps_uuid).
      

2. Obtain the original CPS UUID file:

   • Locate the original CPS server or backup where the CPS UUID file is stored.
   • Copy the original CPS UUID file to the new CPS server using a secure method (e.g., SCP).
      

3. Replace the CPS UUID file:

   • Log in to the new CPS server.
   • Open a terminal or command prompt.
   • Navigate to the directory where the CPS UUID file is located. For example:
     
     cd /etc/VRTSCPS/db/current/
      
   • Replace the current CPS UUID file with the original CPS UUID file using the following command:
     
     cp <patch_to_original_cps_uuid_file> cps_uuid

     Replace with the actual path to the original CPS UUID file.
      

4. Unregister nodes and delete the cluster:

   • Open a terminal or command prompt.
   • Run the following command to unregister the nodes:
     
     cpsunregister -n <node_name>

     Replace with the name of each node in the cluster. Repeat this command for each node.
      

   • Run the following command to delete the cluster:
     
     cpsdelete -c <cluster_name>
     
     Replace with the name of the cluster.
      

5. Add the cluster and nodes back:

   • Open a terminal or command prompt.
   • Run the following command to add the cluster:
     
     cpsadd -c <cluster_name>

     Replace with the name of the cluster.

   • Run the following command to add each node to the cluster:
     
     cpsadd -c <cluster_name> -n <node_name>

     
     Replace with the name of the cluster and with the name of each node. Repeat this command for each node.

6. Verify the CPS UUID in the running configuration:

   • Open a terminal or command prompt.
   • Run the following command to check the CPS UUID in the running configuration:
     
     cat /etc/VRTScps/dbs/current/cps_uuid
     
     Ensure that the CPS UUID matches the original CPS UUID.
      

7. Reboot the nodes:

   • Reboot each node in the cluster to apply the changes.
   • After the reboot, the nodes should read the old CPS UUID from the CPS DB, and fencing should start normally.

By following these steps, you can replace the CPS UUID file for Coordination Point Servers migration and resolve issues related to fencing and cluster node reboots.