[rabbitmq-discuss] Pathological dead queue behavior in a cluster.

[Matthew Sackman]

On Wed, Mar 30, 2011 at 10:23:51PM +0100, Matthew Sackman wrote:
> Actually, if you did a basic.cancel and then a basic.consume then you
> should find you can start consuming from the queue again. That is the
> expected strategy in light of the above notification extension.

Err, the basic.cancel is unnecessary: the client and server both tidy up
any state associated with the subscription when they receive the
notification. You should just need the new basic.consume. However, this
behaviour is certainly specific to 2.4.0 clients and broker: prior to
this, if you have reason to suspect the queue had silently gone away,
you'd have to cancel and then reconsume otherwise at the very least,
you'd have a (very very slow) memory leak.

Matthew