IP Failover – Question about "failback" process – is it automatic?

While reading the IP failover article here as well as Apple’s HA document, I got the impression that failing back to the master server from the failover server is not an automatic process. There seems to be a lot of unplugging/rebooting/counting/plugging/holding breath/sacrificing chickens, etc. Am I mistaken here? If I have a master server go down, for whatever reason, and a failover takes place, then can I not simply power up the master server after fixing it and have a nice seamless failback? Or does it work like some big mousetrap that you have to manually set after any failover event?

Cheers,
Andrew Hurt