Replication copies data changes from one MySQL server (source) to others (replicas) in near real-time. It powers read scaling, high availability — and gives analysts a server they can hammer without touching production.
Key Points
- The source writes changes to a binary log; replicas replay it
- Read scaling: send SELECT-heavy analytics to replicas
- Failover: promote a replica if the source dies
- Asynchronous by default; semi-synchronous adds a durability guarantee
- GTID-based replication simplifies failover and consistency tracking
- Best practice: run BI tools and heavy reports against a replica
.png)