[rabbitmq-discuss] Question on Cluster Restart

Wed Jul 24 10:39:15 BST 2013

On 23/07/13 19:58, Richard Raseley wrote:
> In reading this page:
>
> http://www.rabbitmq.com/clustering.html
>
> It seems that when shutting down a cluster for upgrade you must ensure
> that the last node down is the first node up. However, in previous
> versions of the documentation I recall this being a caveat for if and
> only if you had a mix of disk and RAM nodes. Is this not the case
> anymore (we have all disk-backed nodes).

I don't think that requirement has changed since we added cluster upgrades.

Cheers, Simon

-- 
Simon MacMullen
RabbitMQ, Pivotal