The MDBean which acts as a message consumer in the JMS architecture seems to have died in our JBOSS.
Our JBOSS is running in production.
Basically we observed there is a huge lot of requests pending for the concerned MDBeans to pick up. We observed that the Request dispatcher is correctly sending the data to the concerned queue but then the corresponding MDBean's onMessage method is never invoked. Later we came to know that this consumer or MD Bean does not exist at all.
The concerned MDBean was working fine which all of a sudden seemed to have died as the "listThreadCpuUtilization" method in Server-info of JMX-console does not show the existence of the concerned MDBean or the message consumer.
Now we have lot of pending requests to be processed by this consumer as the pool size of this MDBean was mentioned to be huge. We can restart the jboss to let the things work normally but then it is not allowed.
Some how we need to restart this consumer which seems to have died or non-existant without restarting the jboss. A sample java client accessing these consumers and starting them again will also be handy