Replacing a failed node in a high availability configuration

If one of the nodes in your HA group fails, you can replace it.

About this task

The steps to take to replace a node depend on the scenario:

If you are replacing the failed node with a node with an identical configuration, you can replace the node without disrupting the HA group.
If the new node has a different configuration, then you must delete and then rebuild the HA group. You can first back up the queue managers from the node they are running on, then restore them after you have rebuilt the HA group.

If the replacement node is configured to look like the failed node (same hostname, same IP addresses, and so on), then complete the following steps on the new node:
1. Create an rdqm.ini file that matches the files on the other nodes, and then run the rdqmadm -c command (see Defining the Pacemaker cluster (HA group)).
2. Run the crtmqm -sxs qmanager command to recreate each replicated data queue manager (see Creating an HA RDQM).
If the replacement node has a different configuration to the failed node:
1. If required, back up your queue managers (see Backing up and restoring IBM MQ queue manager data).
2. Delete the replicated data queue managers from the other nodes in the HA group by using the dltmqm command (see Deleting an HA RDQM).
3. Unconfigure the Pacemaker cluster by using the rdqmadm -u command (see Deleting the Pacemaker cluster (HA group)).
4. Reconfigure the Pacemaker cluster, including the information for the new node, by using the rdqmadm -c command (see Defining the Pacemaker cluster (HA group)).
5. If necessary (that is, if you do not have SSH access to the other nodes) run the crtmqm -sxs qmanager command to recreate each replicated data queue manager on the other nodes (see Creating an HA RDQM).
6. Run the crtmqm -sx qmanagercommand to create the queue managers on the replacement node .
7. If required, restore data and configuration to your queue managers (see Backing up and restoring IBM MQ queue manager data).