I have a master and a slave running streaming replication between the two servers. I would like to setup some kind of third machine to play the role of observer node to monitor the replication and do the automatic fail over. However, I can't find any documentation or suggestions on such an implementation. Is it possible to do with 9.2.4 or will I need to rely on some sort of F5 load balancer to handle the auto failover?
