A vSphere administrator is unable to power on a VM in a vSphere HA-enabled cluster. Which vSphere HA feature could cause this?

Datastore Heartbeats are a secondary isolation detection mechanism used to distinguish a truly failed host from one experiencing a network partition, and they do not control whether a specific VM can be powered on.

Admission Control reserves cluster resources to guarantee failover capacity and can block power-on operations when insufficient resources remain, but it is not the feature that explicitly excludes a specific VM from HA restart operations.

When VM Restart Priority is set to Disabled, vSphere HA explicitly excludes that VM from the restart workflow and will not attempt to power it on following a host failure. This configuration opts the VM out of HA protection entirely, so no HA-initiated power-on action will occur. To allow HA to restart the VM, the administrator must assign a priority level - Low, Medium, or High - rather than leaving it Disabled.

VM Monitoring uses VMware Tools heartbeat data to detect guest OS failures and reset unresponsive VMs, but its enabled or disabled state has no effect on whether a VM can be initially powered on.