JMS Broker with WebSphere


We are having any issue where websphere is loosing connections from Broker Queue after sometime.
In our case JMS Broker is the provider and has a Queue setup on top of it and websphere is the consumer of the messages.

Just trying to understand if webMethods JMS Broker is timingout the connection from websphere. Anyone has such an experience.

Any help is appreciated.

Vijay Josyula

Vijay Josyula,

My current project has not experienced any timeout issues using Websphere MDB’s (with durable subscribers) to consume messages from BrokerJMS topics (not queues).

We are using the webMethods naming provider (wmjmsnaming.jar) as the JNDI provider and BrokerJMS as the JMS provider.



In an effort to leverage an existing thread, I’m going to revive this one as it matches my scenario closest.

Our JMS provider is Websphere MQ and a clustered wM 7.1.1 JMS trigger is the client. In a non-clustered environment, the JMS trigger service fires as expected. However, in a clustered env, the JMS trigger services executes on both nodes.

Per the wM_IS_JMS_Client_Developer’s_Guide_7_1_1.pdf page 107 which describes how to configure JMS triggers for load balancing, I think we need to subscribe to the topic as a “durable subscriber”. It also says that the durable subscriber (i.e. wM JMS trigger) must be configured to share state and that this can be done in My webMethods Server.

Has anyone set up a 7.1.1 in a cluster successfully. If so, can you confirm that wM has to be setup as a durable subscriber? Since MQ is our JMS provider, is the “durable subscriber” setup in MQ?