You may also get almost the same availability of OPS by having standby database and having Failover agents (say Veritas Cluster Server) running on different machine.

When one machine fails, the other system will take over the responsibility of the primary server. (There will be a down time of few seconds to few minutes, for the second system to take over)

But again, the Data is residing on the same fileserver (say NetApp) which is accessible by both the servers.

In this configuration, only one machine is accessing the DB at anytime.

I guess the advantage here is low "locence" cost.