Sharding
In plain terms
Splitting a database across multiple machines by partitioning the data.
Going deeper
Sharding distributes rows across servers (shards) based on a key, so each holds a slice of the data. It lets databases scale beyond one machine's limits, at the cost of more complex queries spanning shards.
Under the hood
Shard-key choice is critical to avoid hotspots; cross-shard joins and transactions are hard, and rebalancing on node changes is a major operational concern.
Examples
User data partitioned by user_id, Vitess, MongoDB sharded clusters, Citus.
Think of it like...
Like splitting one overflowing phone book into volumes A–H, I–P, Q–Z, each kept in a different office.