Replication
Leader-follower, multi-leader, sync vs async.
Theory
Replication copies data to multiple nodes for fault tolerance and read scaling. Leader-follower (primary-replica) is the most common pattern: all writes go to the leader, followers apply the same changes.
Synchronous replication waits for follower acknowledgment before confirming a write — no data loss on leader failure but higher latency. Asynchronous replication confirms immediately — faster writes but risk of lost un-replicated data.
Multi-leader replication allows writes on multiple nodes (conflict resolution required). Leaderless quorum systems (Dynamo, Cassandra) use R + W > N for tunable read/write consistency without a single leader bottleneck.
Production rollouts require idempotent automation, peer review, staged apply, and documented rollback — treat changes as production code.
Interviewers want STAR stories linking Replication to measurable outcomes: fewer outages, faster deploys, lower cost, or reduced toil.
Architecture Diagram
Users / clients
|
Replication
|
Core services
|
Data + observabilityExamples
# Replication
# Leader-follower, multi-leader, sync vs async.
# Validate in staging before production rollout.Interview Questions
What problem does Replication solve?
It addresses the core use case described in production architecture — map features to reliability, scale, or velocity outcomes.
Key components of Replication?
Identify inputs, outputs, control plane, data plane, and failure domains — interviewers want structured decomposition.
Common production pitfalls?
Misconfiguration, missing observability, no rollback path, and scaling bottlenecks under peak load.
How do you test changes safely?
Staging parity, canary/gradual rollout, automated health checks, and documented rollback.
Metrics to prove success?
Error rate, latency percentiles, throughput, cost, and toil reduction — pick one primary SLO.
Beginner vs advanced concern?
Beginners focus on setup; advanced teams focus on blast radius, security boundaries, and operability at 10× scale.
Best Practices
- Treat Replication config as code with review and CI validation.
- Define SLOs and dashboards before production cutover.
- Document rollback and ownership for on-call.
- Use least privilege for credentials.
Common Mistakes
- Adopting Replication without measurable success criteria.
- No staging environment mirroring production constraints.
- Missing rollback path during incidents.
- Undocumented on-call expectations.
Trade-off Analysis
Replication improves leader-follower, multi-leader, sync vs async. but adds operational and cognitive complexity — justify with load and team size.
Favor simplicity until metrics (p99 latency, error rate, cost) prove the pattern necessary.
Every redundancy layer trades capital/operational cost for availability — align with explicit SLO targets.
Document accepted inconsistency windows and recovery behavior before production cutover.
Cheat Sheet
Practical Exercises
Stand up Replication locally or in free tier; document commands and failure recovery.
Introduce misconfiguration; practice detection and rollback under time limit.