Sharding

Distributing a large dataset across multiple database instances (shards), each holding a subset of the data. Partition key determines which shard holds each row. Enables horizontal scaling beyond single-node limits.