[rabbitmq-discuss] RabbitMQ network issue (IP change) on cluster

Sun Oct 30 03:26:56 GMT 2011

Hello all.

I have 3 nodes running the cluster.

It works fine at beginning, but when I run /etc/init.d/networking restart
(the command can lead to some node with different IP).

The cluster may not to find the node which change the ip.

For example, after the command, node 2 change the ip.
Now node 1 and node 3 are in cluster and can not find node2, but they still
have the information about the cluster.
Node 2 can not find node 1 and node 3 either.

root at lab002:/home/ooo# rabbitmqctl cluster_status
Cluster status of node rabbit at lab002 ...
[{nodes,[{disc,[rabbit at lab003,rabbit at lab002,rabbit at lab001]}]},
 {running_nodes,[rabbit at lab002]}]
...done.

In this case I need to restart the node 2 to make it find node 1 and node 3.

Is the cause that the erlang VM can not detect the ip change on nodes when
running?

Thanks

Shen, Yu-Teh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rabbitmq.com/pipermail/rabbitmq-discuss/attachments/20111030/efe71025/attachment.htm>