Summary
Active-active runs traffic in multiple regions or clusters at once. Active-passive keeps a standby environment ready for failover.
Interview Points
- Active-active improves availability and latency but increases data consistency complexity.
- Active-passive is simpler but may have slower failover and underused capacity.
- Define RPO, RTO, traffic routing, health checks, and data replication.
- Split-brain and failback are major risks.
- Not every system needs active-active.
2-3 Minute Interview Script
“Active-active means multiple environments serve production traffic at the same time. Active-passive means one environment serves traffic while another waits as a standby.
Active-active can improve latency and availability because users can be routed to nearby healthy regions. The tradeoff is complexity: data replication, conflict resolution, global traffic management, and split-brain risk.
Active-passive is easier to reason about. It can meet many business needs if the RTO and RPO allow some failover time and possible data lag.
In an interview, I would start with recovery requirements. If the product needs very low downtime across regions, active-active may be justified. If not, active-passive is often cheaper and safer.”
Follow-Ups
- How do you prevent split-brain?
- What is failback?