Cannot delete a queue that shows unsynchronized and running but no activity

[V Z]

This may sound like an issue that was already discussed on the forum (with no outcome), but I would like to report it since not everything is the same.

We run 4 node cluster (Rabbit 3.6.2, Erlang 18.1). Management interface stopped responding today (cannot get to the login page; even if login page comes up, login hangs -- have to kill Chrome tab; API calls time out). There was nothing particular in the logs. Rabbit itself seems to have been working fine. We restarted one of the nodes, and management interface came to life on this and remaining nodes. 

However, some queues show unsynchronized (even though there are no messages), running (even though there are no consumers and no bindings besides default), and cannot be deleted or synchronized (trying to do that simply times out). Attached is a snapshot of one of such queues. 

Please see sasl log attached from the master node. 

[Michael Klishin]

It is server#812.

--
You received this message because you are subscribed to the Google Groups "rabbitmq-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rabbitmq-user...@googlegroups.com.
To post to this group, send email to rabbitm...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

<rab...@4b99d38df77b2d3eeaf224cf610d1787-sasl.log>

[V Z]

So, wait for 3.6.3? 

What should we do in the mean time? Restart all nodes? Clean something up? Rebuild cluster from scratch?

I tried to delete the vhost but no luck. Rabbit log is full of perpetual non-stop queue sync attempts.

[Michael Klishin]

https://github.com/rabbitmq/rabbitmq-server/issues/812

Restarting all nodes at once should be sufficient.

--
You received this message because you are subscribed to the Google Groups "rabbitmq-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rabbitmq-user...@googlegroups.com.
To post to this group, send email to rabbitm...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

MK

Staff Software Engineer, Pivotal/RabbitMQ

[Wong Energy]

I have encountered this situation several times, and some queues suddenly became unsynchronized.
You can try this command(replace virtualhost and queue), which can be deleted in most cases.

rabbitmqctl eval 'Q = {resource, <<"VirtualHost_Name">>, queue, <<"QUEUE_NAME">>}, rabbit_amqqueue:internal_delete(Q, <<"CLI">>).'

But there are also some that cannot be deleted, and I don't know why.