How HA failover and giveback in cluster mode is different from 7 Mode.

[MIP_GMT01]

Basically trying to figure out why failover and give back in cluster mode is suppose to be more efficient.

[aborzenkov]

First, takeover after error is the same in both cases. For manual takeover, 7-mode starts serving data after all aggregates were taken online. C-Mode will move aggregate ownership one by one. So the period of unavailability for each aggregate is reduced.

[MIP_GMT01]

As I understand 7-mode doesn't rake ownership for aggregates infect it creates a virtual copy of the failed over partner in the memory and reconnects connections to it.

[aborzenkov]

It takes ownership of disks, blocking partner from accessing them.