I started by explaining that the API server acts as the central point for all interactions with the cluster. I mentioned that it validates and processes requests before storing state in etcd. However, when the interviewer asked about how leader election works in Kubernetes, how etcd handles data consistency, and the role of RAFT consensus in API server operations, I didn’t have strong answers. I tried to discuss general distributed system principles, but it was clear they were expecting a deeper understanding of Kubernetes internals.