Failover clustering serves the goal of high availability. High availability is a nonfunctional requirement for many systems, but it is difficult to transform its benefits into ROI. HA is a kind of insurance: It pays only if you have an incident. The shorter the downtime, the lesser the loss of productivity.