Failover is the process of maintaining uptime for your database. It can be automatic or manually triggered. The enterprise must calculate how much downtime it can tolerate and plan to have less than that. This is sometimes called high availability. The enterprise must also plan for how to handle the loss of a database server. How does it recover? How quickly? What are the sources of failure? How much complexity can your enterprise manage? Will you rely on backup, mirroring, or clustering? Will you use an internal database solution or a third party application?