Unable to consume from queue if the previous consumer did not exit cleanly

If the receiver on a queue do not exit cleanly other clients are blocked to receive messages from that queue.

It seems that there is an clean-up issue after non-clean exit of a consumer that leads to that other consumers are blocked to receive messages from that queue for ~40-60 seconds. It may be considered as a DoS issue.

This applies to all proton clients (qpid-jms, python-proton, qpid-cpp).

1. send a message to a queue
./aac1_sender.java.sh --log-msgs dict --broker localhost:5672 --address jms.queue.interoperability --count 1 --msg-content test
2. receive the message
./aac1_receiver.java.sh --timeout 10 --log-msgs dict --broker localhost:5672 --address jms.queue.interoperability --count 1
3. message received, everything is ok so far
4. run the receiver again
5. as there are no message to be consumed use ^C to end the client (or kill the client)
6. send another message to that queue
./aac1_sender.java.sh --log-msgs dict --broker localhost:5672 --address jms.queue.interoperability --count 1 --msg-content test
7. try to consume that message
clients/aac1_receiver.java.sh --timeout 10 --log-msgs dict --broker localhost:5672 --address jms.queue.interoperability --count 1
8. message can't be consumed
9. run the receiver again, this time with higher time-out
./aac1_receiver.java.sh --timeout 60 --log-msgs dict --broker localhost:5672 --address jms.queue.interoperability --count 1
10. message is consumed after ~40-60 seconds
11. the receiver on that queue was blocked for unreasonable amount of time due to the previous client that doesn't exit cleanly.