Etcd powers Kubernetes and many distributed systems, yet its internals remain a mystery to most engineers. In this advanced, demo-driven session, we’ll open the black box and explore how etcd really works: Raft consensus, WAL and snapshot mechanics, MVCC, and compaction. Through live failure and recovery demos, you’ll see what happens during leader elections, partitions, and performance bottlenecks. You’ll leave with practical patterns for tuning, securing, and operating etcd in real production environments.