VMware vSphere® High Availability (HA) minimizes virtual machine downtime by monitoring hosts, datastores, virtual machines, or applications within virtual machines, then, in the event a failure is detected, restarting virtual machines on alternate hosts or resetting them on the same host.
All active hosts in a cluster (those not in standby mode, maintenance mode, or disconnected) participate in an election to choose the primary host for the cluster; all other hosts become secondary hosts. The primary has a number of responsibilities, including monitoring the state of the hosts in the cluster, protecting the powered-on virtual machines, initiating failover, and reporting cluster health state to vCenter Server. The primary is elected using an algorithm that takes into account various properties of the hosts, such as the number of datastores to which they are connected. Serving in the role of primary will have little or no effect on a host’s performance.
When the primary host can’t communicate with a secondary host over the management network, the primary uses datastore heartbeating to determine the state of that secondary.
In the event of a failure, vSphere HA queries DRS running on vCenter for its recommendation about which hosts the virtual machines should be restarted on. If DRS doesn't supply a recommendation, HA falls back to a simpler algorithm.
For further details about vSphere HA, see vSphere Availability.
