Internally, VSAN will initiate a rebuild of degraded components according to a schedule or in the event of a failure. Should a disk or disk group fail, the components on the affected disks will be immediately rebuilt if there is adequate capacity and enough nodes to comply with the policy decisions applied to the affected objects.

In the absence of a physical fault, VSAN implements rebuild logic according to internal timers. When a node goes offline and exits the cluster, VSAN will take no action for 60 minutes. After one hour has passed, VSAN will assume that the node has exited permanently or semi-permanently and will begin to rebuild any degraded objects to restore fault-tolerance to the cluster. Re-entry of the node will not be forbidden, nor will it cause any problems.

The cluster departure timer is not disabled during maintenance operations. A powered-on/functional host in maintenance mode is in a decommission state, but it is still part of the cluster. Once the host goes down, however, the timer kicks in and the components associated with the host go into an ABSENT state. After 60 minutes elapse, even if the host was in maintenance mode, the rebuild will begin. For this reason, if the maintenance-related outage will be extended (such as during significant or lengthy hardware changes), it is recommended that all data be migrated off in advance as a part of the maintenance mode operation.