An administrator is creating a hybrid vSAN cluster with six ESXi hosts equally distributed on three server racks. How can this cluster be protected from a single rack failure?

Multiple storage controllers per host improve per-host I/O redundancy and throughput but have no effect on where vSAN places data replicas across the cluster, leaving data vulnerable to a rack outage.

Creating multiple disk groups per host increases a single host's cache and capacity tiers but does not influence replica placement across racks or protect against the simultaneous loss of all hosts in one rack.

Additional flash caching devices within a disk group increase cache capacity and read performance within a host but provide no rack-level data placement protection whatsoever.

Creating three fault domains - one per rack - instructs vSAN to treat each rack as a single failure boundary and ensures that replica copies of data objects are placed on hosts in different fault domains. With three fault domains and a default Failures to Tolerate value of 1, vSAN can sustain the complete loss of one rack and still satisfy the storage policy. Without fault domains configured, vSAN may place replicas on hosts within the same rack, leaving data unprotected against a rack-level failure.