[rabbitmq-discuss] best practices for HA?

Wed Sep 10 16:35:04 BST 2014

Hi,

I'm working on an OpenStack controller using RabbitMQ 3.2.4 and I want 
to extend it to a 2-node HA setup.

I'm looking at how to handle the HA for RabbitMQ.

The RabbitMQ docs (at https://www.rabbitmq.com/ha.html) seem to 
recommend active/active with mirrored queues.

On the other hand, the OpenStack high availability guide currently 
recommends using active/standby with the messages stored on a replicated 
filesystem.  Do you know why they would recommend that over the 
active/active model?

Thanks,
Chris