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

  1. Why replicate, and the single-leader model
  2. Topologies: single-leader vs multi-leader vs leaderless
  3. Synchronous, asynchronous and semi-synchronous replication
  4. How changes flow: WAL streaming vs logical replication
  5. Read replicas and query routing in the application
  6. Read-after-write, monotonic reads and consistent prefix
  7. Replication lag: causes, measurement and fixes
  8. Failover, promotion and split brain
  9. Write scaling limits: when replicas stop helping

Previous: Connection Pooling, Lock Contention & Deadlocks · Next: Partitioning, Sharding & Zero-Downtime Migration