[rabbitmq-discuss] No period allowed in queue name in rabbit 1.8.0?

[Nicolás César]

I've updated to 1.8.0 and with the same code I get a 404 NOT_FOUND when i use queue_declare

---------------------
Traceback (most recent call last):
(...)
  File "/usr/local/sbin/test2", line 138, in __init__
    self.chan_in.queue_declare(queue='st.test-o.com_q', durable=True, exclusive=True, auto_delete=False)
  File "/usr/lib/pymodules/python2.5/amqplib/client_0_8/channel.py", line 1296, in queue_declare
    (50, 11),    # Channel.queue_declare_ok
  File "/usr/lib/pymodules/python2.5/amqplib/client_0_8/abstract_channel.py", line 105, in wait
    return amqp_method(self, args)
  File "/usr/lib/pymodules/python2.5/amqplib/client_0_8/channel.py", line 274, in _close
    (class_id, method_id))
amqplib.client_0_8.exceptions.AMQPChannelException: (404, u"NOT_FOUND - no queue 'st.test-o.com_q' in vhost '/'", (50, 10), 'Channel.queue_declare')
----------------------


I've changed st.test-o.com_q  to  st_test-o_com_q and had no problem... periods are not allowed in queue names now?

Nico César
http://blog.nicocesar.com

[Simon MacMullen]

On 28/06/10 18:06, Nicolás César wrote:
<snip>

> I've changed st.test-o.com_q to st_test-o_com_q and had no problem...
> periods are not allowed in queue names now?

No, queue names can still have periods in.

ISTR the only way you can get a not-found when declaring a queue is in a
clustering situation, if the queue exists on a node which is not
running. Does that sound like it could be a possibility?

Cheers, Simon
--
Simon MacMullen
Staff Engineer, RabbitMQ
SpringSource, a division of VMware

_______________________________________________
rabbitmq-discuss mailing list
rabbitmq...@lists.rabbitmq.com
https://lists.rabbitmq.com/cgi-bin/mailman/listinfo/rabbitmq-discuss

[Simon MacMullen]

On 28/06/10 20:17, Nicolás César wrote:
> 2010/6/28 Simon MacMullen <si...@rabbitmq.com <mailto:si...@rabbitmq.com>>

> ISTR the only way you can get a not-found when declaring a queue is
> in a clustering situation, if the queue exists on a node which is
> not running. Does that sound like it could be a possibility?
>
>
>

> Thanks Simon for your quick response. It's a single node. but when i've
> tried to check:
>
> # rabbitmqctl cluster
> Clustering node rabbit@sendsorium with [] ...
> Error: mnesia_unexpectedly_running
>
> what does mnesia_unexpectedly_running means??

It means pretty much what it says. rabbitmqctl cluster is used to set up
a cluster, and should only be used when the rabbit application has been
stopped (which stops mnesia too). See:

http://www.rabbitmq.com/rabbitmqctl.1.man.html#Cluster%20management

The command you're looking for is:

rabbitmqctl status

By the way, it's a good idea to keep the mailing list CCed so that if
other people have the same problem they can find the thread...

[Iver Odin Kvello]

Nicolás César <nico@...> writes:

>     (class_id, method_id))amqplib.client_0_8.exceptions.AMQPChannelException:

(404, u"NOT_FOUND - no queue 'st.test-o.com_q' in vhost '/'", (50, 10),

'Channel.queue_declare')----------------------I've changed st.test-o.com_q  to 

st_test-o_com_q and had no problem... periods are not allowed in queue names
now? Nico Césarhttp://blog.nicocesar.com
>

We just had the same problem, which apparently was due to having some residues
of the old (durable) queues from the previous install present; in particular,
when using

rabbitmqctl list_bindings

the old queues were visible, whereas

rabbitmqctl list_queues

showed no queues present. It was then possible to declare new queues, but not
queues with the same name as the ones that existed before.

The problem was fixable by
rabbitmqctl stop_app
rabbitmqctl reset
rabbitmqctl start_app

after which the old bindings were gone and the queueus could be declared again.

Regards,
Iver Odin Kvello