Duplicate Message Processing by JMS Triggers with a SharedQueue Durable
In some situations, a concurrent JMS trigger that uses a SharedQueue durable may perform duplicate processing of messages. This occurs when the Max execution threads for the trigger is set to a value greater than 1 and JMS trigger message processing is slow. The message goes into a pending state and is then rolled back to Universal Messaging. The trigger then receives, and possibly processes, the message again.
To resolve this issue, do one or more of the following:
Change the durable type on the connection factory in
Universal Messaging to Shared.
Set the watt.server.jms.trigger.pooledConsumer.timeout server configuration parameter to a higher value. Setting a high value for watt.server.jms.trigger.pooledConsumer.timeout results in the JMS session and messageConsumer remaining open for a long period of time.