A well-configured HA cluster will have enough free resources to restart all the business-critical VMs running on the hosts, in the event of host failures. This amount of free resources is referred to as the failover capacity.
Failover capacity determines the number of the ESXi hosts that can fail in an HA cluster, and still leave enough resources to support all the powered-on VMs. We can use admission control to control and monitor the failover capacity.
There are three admission control methods (policies):
Note that admission control ...