Here's a simplified diagram of my broker network: <http://activemq.2283324.n4.nabble.com/file/n4729601/Broker_Network.png>
Application 1 and Application 2 both both use messaging internally. Their messages do not need to be distributed across the broker network. The applications consume messages from the clients that all connect to Broker 3. However, processing of those messages is relatively slow. Therefore, Broker 1 and Broker 2 are configured with destination policies for the external destinations so they do not consume 100% of the brokers' resources, and we enable flow control so the clients will slow down production of those messages. If I understand you correctly, we must also configure destination policies for those external destinations on Broker 3. Otherwise, when Application 1 pushes back, Broker 3 would use 100% of its resources for that destination, right? Furthermore, I cannot just copy the destination policies: I need to determine what percentage of Broker 3's resources will be used by Application 1 and Application 2 messages. Does that sound right? -- View this message in context: http://activemq.2283324.n4.nabble.com/Network-of-brokers-and-destination-policies-tp4729188p4729601.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.