Databases
Advanced
What is Database Sharding?
A horizontal scaling strategy that distributes data across multiple database servers based on a partition key.
Sharding splits a database into shards, each on a separate server. A shard key (e.g., user_id) determines which shard stores each row. This distributes both data and query load, enabling horizontal scaling beyond a single server's capacity.
Challenges include cross-shard queries (joins across shards), rebalancing shards as data grows, maintaining consistency, and choosing an effective shard key. Some databases support native sharding (MongoDB, CockroachDB). Often avoided until truly needed due to complexity.
