[rabbitmq-discuss] Join_Cluster Question

[Tim Watson]

Hi Richard,

On 7 Feb 2013, at 03:53, Richard Raseley wrote:

> When creating a cluster using "rabbitmqctl join_cluster rabbit at hostname" can you use an IP address or FQDN instead of a hostname after the @?
> 

Well, Erlang does allow you to use an IP address after the node name, but whilst running `rabbitmqctl status -n rabbit@<hostname>` might work for me, using the IP address instead of the hostname doesn't:

t4 at iske:rabbitmq-server $ ./scripts/rabbitmqctl status -n rabbit at 192.168.1.75
Status of node 'rabbit at 192.168.1.75' ...
Error: unable to connect to node 'rabbit at 192.168.1.75': nodedown

=ERROR REPORT==== 7-Feb-2013::10:45:22 ===
** System NOT running to use fully qualified hostnames **
** Hostname 192.168.1.75 is illegal **


Erlang seems to think that rabbit@<ip-address> is a fully qualified hostname, whilst rabbit@<hostname> is fine:

t4 at iske:rabbitmq-server $ ./scripts/rabbitmqctl status -n rabbit at yddraig
Status of node rabbit at yddraig ...
......


Rabbit's start scripts launch the Erlang runtime using -sname (see http://www.erlang.org/doc/man/erl.html for details) so an FQDN isn't going to work here.

> I am not sure if my issues is being caused by lack of name resolution or a malformed command and the documentation (found here: http://www.rabbitmq.com/man/rabbitmqctl.1.man.html) didn't seem to specify what a valid value was.
> 

The command is fine - if it wasn't you'd get an appropriate error message saying so. Lack of name resolution is almost certainly the problem here.

Cheers,
Tim