TL;DR — Redis Cluster

Design a distributed in-memory key-value store similar to Redis Cluster. The system must partition data across multiple nodes using hash slots, replicate data for fault tolerance, and automatically detect and recover from node failures using a gossip-based protocol. Clients should be able to read and write keys without knowing the cluster topology. Key features: Key-value operations: GET, SET, DEL, EXPIRE with sub-millisecond latency. Data partitioning: 16,384 hash slots distributed across master nodes.