Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?


Adding Nodes to a Cluster Overview


Adding any type of node to an existing cluster is very similar to the initial installation. Adding SBR nodes or a management node is easier than adding data nodes because the first two types of nodes do not directly support the database. Adding data nodes requires taking the existing data nodes offline and rebuilding the session database.

In general, to add new nodes to an existing cluster:

  1. Plan the integration of the new nodes by filling out an expansion node worksheet. See Table 10.

  2. Work through the tasks and prerequisites in Preparing for a Steel-Belted Radius Carrier Installation to ensure that the new servers meet all SSR server requirements.

  3. On a management node in the existing cluster, build upon the existing cluster configuration to generate updated cluster definition files that outline the new cluster topology.

    Use the entries from Table 10 to provide the names, IP addresses, and node IDs of the new servers.

  4. Transfer a copy of the tar file containing the new cluster definition files to each existing node and each new node.

  5. Install SBR Carrier software on the new nodes, and configure the existing cluster nodes with new topology.

  6. Restart the cluster so all nodes come online using the new cluster definition files.

The difference between adding SBR Carrier nodes or a management node and data nodes manifests itself in the way Steps 5 and 6 are implemented.

  • When installing SBR Carrier nodes or a management node, each node is reconfigured and restarted, one at a time. Because only one node at a time is out of service, the cluster’s redundancy protects all connections.

  • Adding data nodes requires taking the existing data nodes offline and rebuilding the SSR database, so all nodes are reconfigured, then all nodes are restarted as a cluster.


The procedures in this chapter describe how to add an SBR node, a management node, or data nodes to an existing cluster. These procedures assume a basic cluster configuration of two (sm) nodes and two (d) nodes. Because your cluster configuration may differ from this, review these procedures in their entirety before starting the procedure. If your configuration is different, you may need to make changes to these procedures. If you have to modify these procedures, see When and How to Restart Session State Register Nodes, Hosts, and Clusters, to determine the proper order for starting and stopping the various node types in your cluster.