specific question about cluster SPP and his service exposed

1. How should a service be exposed to high reliability (I imagine that a clustered service has associated fqdn and ip)?

2. Does an automatic service failover mechanism exist or has it been implemented? In case not can we request an enhancement request?