[rabbitmq-discuss] Upgrading 2.7.1 to 2.8.0 in a cluster // newcomer question

Mon Mar 19 16:52:55 GMT 2012

Relative newcomer to RabbitMQ trying to go from 2.7.1 to 2.8.0

We have 3 servers in a development cluster

srv1 Disc
srv2 RAM
srv3 RAM

I brought down the two RAM nodes first, then the disc node.  We're running the RPM version on Centos/RHEL, so I updated using rpm -U on the disc node.  All is well on that node and it's up and reporting 2.8.0, etc from the management interface.

Started srv2 with a hold on the service start, updated it with rpm -U also, then started it and it fails with an error in startup_log

****

Cluster upgrade needed but this is a ram node.
Please first start the last disc node to shut down.

****

Same for srv3.

The disc node is running and the status (from srv1) is:

Cluster status of node 'rabbit at mq-clustera-srv1' ...
[{nodes,[{disc,['rabbit at mq-clustera-srv1']},
         {ram,['rabbit at mq-clustera-srv3','rabbit at mq-clustera-srv2']}]},
{running_nodes,['rabbit at mq-clustera-srv1']}]
...done.

From srv2, I think the cookie is fine because here's the status from there:

[root at mq-clustera-srv2 rabbitmq]# rabbitmqctl -n rabbit at mq-clustera-srv1 cluster_status
Cluster status of node 'rabbit at mq-clustera-srv1' ...
[{nodes,[{disc,['rabbit at mq-clustera-srv1']},
         {ram,['rabbit at mq-clustera-srv3','rabbit at mq-clustera-srv2']}]},
{running_nodes,['rabbit at mq-clustera-srv1']}]
...done.

Thoughts?  Any pointers?

Thanks

EKG

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rabbitmq.com/pipermail/rabbitmq-discuss/attachments/20120319/2bb17a14/attachment.htm>