Final answer:
For failover, a cluster must have at least 2 nodes to maintain functionality; this allows for redundancy so if one fails, the other can take over (option b). The cluster cannot sustain the loss of 3 or more nodes as it would compromise its operation during failover.
Step-by-step explanation:
For failover purposes, a cluster must have a minimum of how many nodes to maintain functionality and continue operating even if it loses some of its nodes. The answer is b) 2 nodes. With only a single node, there wouldn't be any redundancy, and the loss of that node would result in the failure of the cluster. To ensure functionality during a failover event, at least two nodes are required so that if one fails, the other can take over. However, the cluster can't lose three or more nodes, as that would compromise its ability to maintain redundancy or to operate correctly during a failover.
A two-node cluster is the minimum configuration that allows for redundancy, ensuring that one node can always serve as a backup if the other goes down. This setup is critical for maintaining system availability and ensuring that services remain uninterrupted. When designing a cluster, these considerations are crucial to achieving the desired level of reliability and uninterrupted service.
Hence, the answer is option b.