gistfile1.txt

I am currently in the design phase of a high availability RabbitMQ cluster. Our initial deployment is in a cloud computing environment, and, as such, I have been tasked with determining how RabbitMQ will react to network segmentation (should it occur), and how we we should respond to a network segmentation event. Are there any good resources or documentation for this right now?

At the moment I am setting up a low volume, 3 node cluster with 10 queues. All queues are mirrored to exact 2 (All in this case, but we will add nodes later I am sure). All queues, exchanges and messages are durable. 

I have access to a load balancer if I would like to use one. 

Any suggestion or references would be great. If none exist, I will start asking more specific scenario questions to attempt to generate a general guideline for everyone.