Encountering JMSOrderException When One of the Managed Servers in the Cluster is Not Running (Doc ID 2646661.1)

Last updated on NOVEMBER 17, 2023

Applies to:

Symptoms

The JMS Client Producer application encountering JMSOrderException when one of the Managed Servers in the Cluster is not running. This is happening when using Unit Of Order with a Uniform Distributed Queue. 

Use Case

1. Create a WebLogic domain having two managed servers in a cluster.

2. Create two JMSServers (JMSServer1 and JMSServer2) targeting to one managed server each.

3. Create a JMS Module and target it to the cluster of two managed servers.

4. Create a connection factory within the JMS Module created in above step with JNDI as "confuuo".

5. Click on the above Connection Factory==>Default Delivery and select System-generated for Default Unit-of-Order for Producer.

6. Create a Uniform Distributed Queue with JNDI as "disquota" within the above JMS Module.

7. Target the Connection Factory and the Uniform Distribute Queue using a Sub Deployment to JMSServer1 and JMSServer2.

8. Run the attached DisQueueSend.java. It will put 48 messages on the Uniform Distribute Queue with JNDI as "disquota".

9. Shutdown one of the managed servers in the cluster.

10. Run DisQueueSend.java again and JMSOrderException will be raised with following stack trace:

Changes

 None

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.