Oracle Replication for Failover Site

I understand from reading the Oracle Advanced Replication documentation that a failover site can be implemented using a multimaster configuration. On our project, we would like to do this using two of the four servers on a LAN. The part I don't understand is if the primary site becomes unavailable (say the whole server fails) how are the application SQL calls re-routed so that they select/update the data in the failover site. Is this taken care of automatically? If so, what process does this?

-------------------==== Posted via Deja News ====-----------------------
      http://www.dejanews.com/     Search, Read, Post to Usenet