Replication, Read Replicas & Replication Lag
Single-leader, multi-leader and leaderless replication; sync vs async vs semi-sync; WAL streaming vs logical replication; read replicas and routing; read-your-writes and monotonic reads; failover, split brain; lag causes and fixes; why replicas do not scale writes.
About 105 minutes · 15 exam questions · 9 interview questions · 2 code labs
What you will learn
- Why replicate, and the single-leader model
- Topologies: single-leader vs multi-leader vs leaderless
- Synchronous, asynchronous and semi-synchronous replication
- How changes flow: WAL streaming vs logical replication
- Read replicas and query routing in the application
- Read-after-write, monotonic reads and consistent prefix
- Replication lag: causes, measurement and fixes
- Failover, promotion and split brain
- Write scaling limits: when replicas stop helping
Previous: Connection Pooling, Lock Contention & Deadlocks · Next: Partitioning, Sharding & Zero-Downtime Migration