Mysql Fabric is eager to do failover
I've leanrt that Mysql Fabric is really handy in terms of decrease the pain when sysadmins try to create or enable High Availability on Mysql Servers. I'd like to mention a few configuration values that I think are worth to have in mind, specially in environments when the load in your database servers is high enough to making you worry, those values are the ones inside the [failure_tracking] group. Because, in order to detect failures in the cluster and promote the most updated slave to be the new shiny master, some elements should be accomplished first. The formula sounds like this: "let us wait detection_timeout seconds among the trials detections during a detection_interval ", after that period of time. Then if detections has reached the max, then Fabric will promote a slave to master. However, in some environments maybe is worth to touch those values and increase them, for instance, maybe this is acceptable in your cluster, "if we detect 4 times wit