[rabbitmq-discuss] rabbitmq nodedown error

[Matthias Radestock]

On 07/12/12 07:21, pi wrote:
> I tried to stop and start rabbitmq server but keep getting the nodedown
> error as shown below. [...]
> me at myserver:~$ <mailto:me at myserver:~$> sudo rabbitmqctl stop
> Stopping and halting node rabbit at myserver <mailto:rabbit at myserver> ...
> Error: unable to connect to node rabbit at myserver
> <mailto:rabbit at myserver>: nodedown
>
> DIAGNOSTICS
> ===========
>
> nodes in question: [rabbit at myserver]
>
> hosts, their running nodes and ports:
> - rhine: [{rabbitmqctl14960,54442}]
>
> current node details:
> - node name: rabbitmqctl14960 at myserver <mailto:rabbitmqctl14960 at myserver>
> - home dir: /var/lib/rabbitmq
> - cookie hash: lvB7saJGHhQwcaQeveU/WA==

Are you sure that rabbit is in fact running?

> I read that we should synchronise Erlang Cookies but how should that be
> done?

There is no need for that if you are on a single machine and 
consistently use the same set of commands to start/stop the server.

> running the line below shows cookie that's different from cookie hash,
> does it mean that's the cause?
> sudo cat .erlang.cookie
> GBPITQVDFNVGPPZTDYSS

The diagnostics display the base64 encoded md5 hash of the cookie, 
rather than the cookie itself. The same information is recorded in the 
server's startup log, and the two ought to match.

Regards,

Matthias.