[Configure and Utilize Nutanix Unified Storage] An administrator would like to protect an object store from a single node or two-drive failure. What are the requirements for enabling this level of resiliency on a newly-deployed object store?

Protecting Nutanix Objects against the simultaneous loss of either a single node or any two drives requires Replication-Factor-3 (RF3). RF3 keeps three copies of every data chunk, so an object remains available after a complete node outage or two-disk failure. Nutanix Objects will only allow RF3 to be enabled on a cluster that contains at least four physical nodes, because three full copies plus fault-tolerant placement domains are required.

A – RF2 on three nodes survives only one drive OR one node loss, not both. B – Erasure Coding (EC-X) in Objects is capacity-focused and still needs RF2/RF3 underneath; it alone offers no node-level protection. D – RF2 with parity requires fewer copies but still tolerates only a single component failure.

1. Nutanix Objects Administration Guide 3.6, “Replication Factors,” p.26 – “RF3 requires a minimum four-node cluster and tolerates one node or two-disk failure.”

2. Nutanix AOS Data Protection & Resiliency Guide 6.5, §2.3.1 “RF2 vs RF3,” pp.14-15.

3. Nutanix Support KB 7591, “Objects resiliency requirements,” para. 3.

4. Massachusetts Institute of Technology, 6.033 course notes, “Triple-replication fault tolerance,” Example 2 (conceptual background).