The JMS policy based client-side clustering feature in webMethods Broker helps you scale your messaging system. When you implement JMS policy based clustering, you configure a group of two or more Brokers in a cluster to maximize message distribution from the publishing applications to the subscribing applications.

When a Broker comes online or goes offline in a cluster, the webMethods API for JMS throws a connection exception. The JMS client application must catch the connection exception and reconnect. Failover is achieved when the webMethods API for JMS switches the publishing task to another Broker in the cluster based on the policy.

The chapter explains the underlying clustering architecture that supports load balancing for webMethods clients for JMS, lists the most common usage scenarios for these features, and details the steps required for their configuration.