JMS Reconnection fails due to inconsistent connector lifecycle state

Details

Type: Bug
Status: Closed
Priority: Critical
Resolution: Fixed
Affects Version/s: 3.1.2
Fix Version/s: 3.1.3 (EE only), 3.2.0
Component/s: Transport: JMS
Labels:
None

User impact:
High
Log Output:

Hide
testing with ActiveMQ:

ERROR 2011-08-26 12:47:44,052 [MuleServer.2] org.mule.transport.jms.activemq.ActiveMQJmsConnector: org.mule.api.lifecycle.CreateException: Failed to schedule work with the Work manager
ERROR 2011-08-26 12:48:12,048 [MuleServer.2] org.mule.retry.notifiers.ConnectNotifier: Failed to connect/reconnect: ActiveMQJmsConnector

Show
testing with ActiveMQ: ERROR 2011-08-26 12:47:44,052 [MuleServer.2] org.mule.transport.jms.activemq.ActiveMQJmsConnector: org.mule.api.lifecycle.CreateException: Failed to schedule work with the Work manager ERROR 2011-08-26 12:48:12,048 [MuleServer.2] org.mule.retry.notifiers.ConnectNotifier: Failed to connect/reconnect: ActiveMQJmsConnector
Similar Issues:
None

Description

When the jms broker is ungracefully shut down, during the connector stop phase a JMSException is thrown that then is wrapped as a StopException that leaves the connector in "start" state.
Once the broker is up again and the connector is reconnected, is not being started again as its status is already "start". This is causing NPEs because the halfway executed stop process managed to close the scheduler.