A DR administrator must protect production VMs between two Nutanix clusters. The business can tolerate losing up to 15 minutes of writes, and the WAN latency between sites rules out synchronous replication. Which replication mode should be selected?

Nutanix offers three remote replication modes, and the RPO decides between them. Asynchronous replication creates recovery points on a schedule whose shortest interval is one hour, so it cannot meet a 15-minute data-loss limit. NearSync replication covers RPOs from 1 to 15 minutes; it uses lightweight snapshots built from OpLog markers and ships them to the remote cluster at the chosen interval, so it still sends periodic recovery points rather than mirroring every write. That makes it the right fit for a 15-minute RPO over a constrained WAN, provided the link can carry the change rate. Configuring a snapshot frequency of 15 minutes or less automatically engages NearSync. Metro Availability writes synchronously to both sites for a zero RPO, which requires low round-trip latency (5 ms or less) and adds latency to every write, so it is excessive here and impractical over a constrained WAN. A local snapshot policy keeps recovery points on the protected cluster only, which does nothing if that site is lost. Exam caveat: memorize the bands: Metro and synchronous at zero, NearSync from 1 to 15 minutes, asynchronous from one hour upward. Operational check: confirm the protection policy shows NearSync with the 15-minute schedule, check the WAN bandwidth against the change rate, and verify the latest recovery point age at the remote site.