If a high-availability storage cluster loses a majority of its nodes, what typically happens to write operations?

• A. The cluster automatically retries writes on all remaining nodes
• B. The cluster reroutes all I/O to a single node
• C. Writes are blocked until a majority is available
• D. The cluster initiates an automatic failover to a remote site

Answer: (C)

In many high-availability storage clusters, writes require a majority of nodes to agree (a quorum) before they can be committed. This ensures data consistency across the cluster even if some nodes are down or partitions occur. If enough nodes fail so that a majority cannot be formed, the system cannot safely acknowledge or commit writes, so write operations are blocked until a majority is available again or the cluster reconfigures. This prevents conflicting or lost data that could arise from proceeding without quorum. Other approaches, like retrying on the remaining nodes, routing I/O to a single node, or failing over to a remote site without quorum, could lead to inconsistency or data loss, which is why they aren’t the chosen behavior.