RP4VM disaster case

hello,

we use RP4VM 5.3 and all works well. Ich have a few questions in case of disaster:

We have 2 Sites. Each Site has its own ESX Cluster and Storage System. Both Sites use the same vCenter hosted on a third Site.

1. What would happen if the primary Site goes complete down and the VMs and RPAs on the productive Site are not reachable? Can I nevertheless failover?
2. What would happen if the primary Site is repaired and comes up again? Would RP4VM declare them due to failover as copy?
3. If the primary site goes down. Is there a way to failover all protected vms with simply one task and without the testing procedure? Like a disaster button?

Best reguards,

Matthias