[rabbitmq-discuss] RabbitMQ clustering and node failure recovery strategies

[Dave Greggory]

When a node in a cluster goes down hard (let's say HDD failure), how would you 
recommend recovering from that situation (since the durable queues cannot be 
re-created in another node)? Can we bring up a new VM with a new RabbitMQ node 
installed with the same hostname and Rabbit nodename and have it join the old 
cluster? Would that work? Or would it fail because the new node had an empty 
mnesia db and did not have the durable queues in it that the other nodes in the 
clustered expected?