Why are some virtual machines orphaned after rebooting a High Availability (HA) enabled host?

When an HA-enabled host reboots, HA evaluates each VM's restart priority and restarts eligible VMs on surviving hosts. VMs with HA restart set to Disabled are explicitly skipped by this process, meaning they are not powered on anywhere during the outage. When the rebooted host rejoins the cluster, vCenter's inventory still references these VMs but no running instance exists on any host, causing them to appear as orphaned virtual machines.

A recent vMotion that did not persist would leave a registration inconsistency on a specific host, but this scenario is unrelated to the HA reboot event that triggers the orphan state.

Failed storage would make VM disk files inaccessible and could cause VMs to fail, but this would affect all VMs using that storage - not selectively the subset that becomes orphaned after a host reboot.

A host exiting maintenance mode causes DRS to potentially rebalance workloads across the cluster, but this process does not result in virtual machines becoming orphaned.