Nutanix handles node failures by immediately beginning the rebuild process. When two nodes fail within a short time frame, Nutanix automatically initiates the data replication and rebuild process to ensure data integrity and availability. The system leverages the remaining capacity and resources in the cluster to redistribute and rebuild the data from the failed nodes, minimizing the risk of data loss and maintaining cluster resilience.
References:
* Nutanix Bible

Current Configuration:
* The customer has a 3-node cluster configured for RF2 with 100TB of raw disk space.
Erasure Coding Requirements:
* Erasure Coding (EC-X) improves storage efficiency by using a parity-based method to store data.
* For EC-X to be enabled, a minimum of 4 nodes is required in the cluster.
Explanation:
* RF2 (Replication Factor 2): Requires data to be replicated across at least two nodes.
* Erasure Coding: Requires additional nodes beyond the replication factor to store parity information. In this case, a minimum of 4 nodes is needed to implement EC-X.
Required Change:
* Add 1 Node: Adding one more node to the existing 3-node cluster will meet the minimum requirement of 4 nodes, allowing the customer to enable Erasure Coding.
Outcome:
* By adding an additional node, the cluster will be able to utilize Erasure Coding, thereby gaining additional usable space.