Difference between master-slave and master-master replication?

Here’s a clear comparison of Master-Slave vs Master-Master replication in MySQL:

🔹 Master-Slave Replication

• Definition: One server (master) handles all write operations, and the other server(s) (slaves) replicate data from the master.
• Writes: Only allowed on the master.
• Reads: Usually performed on slaves to reduce load on the master.
• Use Case: Load balancing (read scaling), backups, reporting.
• Failover: If the master fails, a slave needs to be promoted manually (or via automation) to act as master.
• Data Flow: One-directional (Master ➝ Slave).
• Conflict Handling: No conflict (since only master handles writes).

🔹 Master-Master Replication

• Definition: Two (or more) servers act as masters, and replicate data to each other.
• Writes: Can be performed on both masters.
• Reads: Can be done on both masters (or additional slaves).
• Use Case: High availability, failover without downtime, distributed writes.
• Failover: Easier, because if one master goes down, the other can take over immediately.
• Data Flow: Bi-directional (Master ⇆ Master).
• Conflict Handling: Possible conflicts if the same row is updated on both masters at the same time (requires conflict resolution strategies).

🔑 Key Differences Table

👉 In short:

• Master-Slave = simpler, safe, mainly for read-heavy systems.
• Master-Master = more complex, better for high availability and distributed writes, but needs careful conflict handling.