Replace a Single Server or Cluster
Caution | Because this process is designed to work as a server replacement, you must perform it in the live environment. Cisco does not recommend doing this process on a "dead net" because a duplication of entire network environment is required, which is highly risky. |
This section provides checklists of the steps that are required to replace a single server or a cluster:
- If you are replacing an entire cluster, replace the servers in the order that is described in the following steps.
- If you are replacing a publisher node in a cluster, or a single server that is not part of a cluster, go to the applicable procedure.
- If you are replacing one or more subscriber nodes or dedicated TFTP servers, go to the applicable procedure.
Replace Publisher Node
Complete the following tasks to replace the Cisco Unified Communications Manager publisher server. If you are replacing a single server that is not part of a cluster, follow this procedure to replace your server.
Replace Subscriber or Dedicated TFTP Server
Complete the following tasks to replace a subscriber node or dedicated TFTP server (or another server type that is not a Cisco Unified Communications Manager server). A dedicated TFTP server is a node with Cisco Unified Communications Manager installed but with the Cisco Unified Communications Manager service disabled. The TFTP service runs as a dedicated TFTP server for the cluster.
If you are also replacing the publisher node, you must replace it before replacing or reinstalling any other nodes. If the cluster uses backup subscriber nodes, replace or reinstall all backup subscriber nodes before replacing or reinstalling primary subscriber nodes.
You can replace all backup subscriber nodes at the same time if this does not cause outages or oversubscription of primary subscriber nodes. You can replace all primary subscriber nodes at the same time if this does not cause outages or oversubscription of backup subscriber nodes.
Follow the references in the For More Information column to get more information about a step.
Replace MCS Hardware with Virtual Machines
You can migrate an existing cluster of Unified Communication Manager nodes that are running on MCS hardware to a new virtualized cluster using the Cisco Prime Collaboration Deployment application. For information, see the Cisco Prime Collaboration Deployment Administration Guide.
Replace a Virtual Machine and Migrate the Cluster
You can migrate an existing cluster of virtualized Unified Communication Manager nodes to a new virtualized cluster using the Cisco Prime Collaboration Deployment application. For information, see the Cisco Prime Collaboration Deployment Administration Guide.