Creating and maintaining duplicate copies of data across multiple locations or servers for reliability and performance. Like having backup copies of important files in different buildings.
A global application replicates user data across data centers in different continents so users get fast access no matter where they are, and data is safe even if one data center fails.
All major clouds support data replication, but the mechanism depends on the data type. Databases commonly use read replicas, multi-region clusters, or log shipping (e.g., Data Guard). Object storage uses bucket/object replication or geo-redundancy. The main differences are replication scope (single-region vs cross-region), consistency model (strong vs eventual), and whether replication is synchronous (lower data loss, higher latency) or asynchronous (lower latency, possible lag).