[rabbitmq-discuss] Failure starting rabbitMQ broker

Simon MacMullen simon at rabbitmq.com
Thu Mar 15 11:03:26 GMT 2012

On 15/03/12 01:43, AlvinRMQ wrote:
> Hi Simon,
> It failed to restart again today. And the log file appeared normal apart
> from a warning mesage just prior to restarting the broker:
> =WARNING REPORT==== 15-Mar-2012::01:09:34 ===
> exception on TCP connection<0.16970.16>  from
> connection_closed_abruptly
> I would think this is from one of the client machines shutting down.

Yes, that's just a client closing the socket without the AMQP close 

Did anything show up in the sasl log?

> Also, in which order should I restart the cluster of 2 nodes? Disc node or
> the ram node first?

You should start the disc node first, the RAM node won't know anything 
if it's started first!

In 2.7.1 starting a RAM node first will give you an empty broker which 
the disc node will then not cluster with. In 2.8.0 RAM nodes refuse to 
start first.

Cheers, Simon

Simon MacMullen
RabbitMQ, VMware

More information about the rabbitmq-discuss mailing list