Related Documentation : Ehcache Replication Guide : Replicated Caching Using JMS : Configuring Clients for Message Queue Reliability
Configuring Clients for Message Queue Reliability
Ehcache replication and cache loading is designed to gracefully degrade if the message queue infrastructure stops. Replicates and loads will fail. But when the message queue comes back, these operations will start up again.
For this to work, the ConnectionFactory used with the specific message queue needs to be configured correctly. For example, with Open MQ, reconnection is configured as follows:
*imqReconnect='true' - without this reconnect will not happen
*imqPingInterval='5' - Consumers will not reconnect until they notice the connection is down. The ping interval
*does this. The default is 30. Set it lower if you want the Ehcache cluster to reform more quickly.
*Finally, unlimited retry attempts are recommended. This is also the default.
For greater reliability consider using a message queue cluster. Most message queues support clustering. The cluster configuration is once again placed in the ConnectionFactory configuration.
Copyright © 2010-2015 Software AG, Darmstadt, Germany.

Product Logo |   Feedback