[rabbitmq-discuss] auto-delete queue/exchanges and cluster synchronization

[mouad ben]

On Tue, May 13, 2014 at 3:44 PM, Michael Klishin <mklishin at gopivotal.com>
 wrote:

> On 13 May 2014 at 17:28:06, mouad ben (mouadino at gmail.com) wrote:
> > > But i think that this also a problem in RabbitMQ side, basically
> > i believe a more sane behavior will be for RabbitMQ to ignore **old**
> > delete if a newest exchange's declare was sent, instead of threading
> > this later as a no op.
> >
> > Does my analyze above make sense !?
>
> I personally think that systems that cannot tolerate the current behaviour
> should not use auto-delete exchanges.
>
> Is this realistic for OpenStack? Is it only auto-delete exchanges that
> cause you practical issues or auto-delete queues, too? If the latter,
> have you considered exclusive queues (note that they cannot be mirrored).
>

Yes it's realistic for OpenStack and i have already filled a bug report
there (https://bugs.launchpad.net/neutron/+bug/1318721), The problem only
happen with auto-delete exchanges not the queues and i believe this is b/c
the queues wasn't mirrored b/c we were relaying on x-ha-policy argument
that do nothing in RabbitMQ 3.*, a misconfiguration from our side sadly,
But i still don't understand why this is the case, why in case of mirrored
queues the problem doesn't happen !


On Tue, May 13, 2014 at 3:44 PM, Michael Klishin <mklishin at gopivotal.com>wrote:

> On 13 May 2014 at 17:28:06, mouad ben (mouadino at gmail.com) wrote:
> > > But i think that this also a problem in RabbitMQ side, basically
> > i believe a more sane behavior will be for RabbitMQ to ignore **old**
> > delete if a newest exchange's declare was sent, instead of threading
> > this later as a no op.
> >
> > Does my analyze above make sense !?
>
> I personally think that systems that cannot tolerate the current behaviour
> should not use auto-delete exchanges.
>
> Is this realistic for OpenStack? Is it only auto-delete exchanges that
> cause you practical issues or auto-delete queues, too? If the latter,
> have you considered exclusive queues (note that they cannot be mirrored).
> --
> MK
>
> Software Engineer, Pivotal/RabbitMQ
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rabbitmq.com/pipermail/rabbitmq-discuss/attachments/20140516/8a30631c/attachment.html>