An external cluster enables you to distribute your processes among multiple Integration Servers without fully implementing an Integration Server cluster. In this case, although the Integration Server nodes are not truly clustered, they must still connect to webMethods Broker (deprecated) using the same client prefix.

Because the Integration Server nodes are not truly clustered, the Process Engine on each node must be configured to behave as if it was operating in a clustered environment. This ensures that each Process Engine executes processes in compliance with cluster requirements (locking step context, for example).

This is done by enabling the External Cluster setting for each Process Engine, as described in Configuring Process Engine Settings. When this setting is enabled, the Process Engine behaves as if it was in a Integration Server cluster.

Each Integration Server must communicate with the webMethods Broker (deprecated) using the same client prefix. As in a standard cluster, you must configure each Integration Server node to use a single external RDBMS containing a Process Engine database.