[rabbitmq-discuss] Scaling Throughput with Highly Available Queues

[Matthias Radestock]

On 19/03/13 21:16, Richard Raseley wrote:
> Sorry, I meant to say the master queue process. Can you comment on my
> previous question assuming that correction?

Yes, all queue operations go via the queue's master.

However, I fail to see what relevance that has to your question about 
scaling throughput via either adding nodes to a single cluster or 
creating new clusters. In either case you are only going to get 
additional scalability at the queue level by having more than one queue 
and distributing those queues them over the available nodes.

Matthias.