Disaster Recovery on AWS: RTO, RPO, and the Four Strategies

Disaster recovery questions give you recovery objectives and a cost sensitivity, then ask which strategy fits. The two metrics: RTO (how fast must service be restored) and RPO (how much data loss is acceptable, measured as the age of the newest recovery point).

The four strategies

• Backup & restore — cheapest, slowest. Snapshots/backups to another Region; rebuild on disaster. RTO: hours+; RPO: backup interval.
• Pilot light — core data layer (databases, critical storage) continuously replicated; compute provisioned at failover. RTO: tens of minutes to hours.
• Warm standby — a scaled-down but functional copy always running in the DR Region; scale up on failover. RTO: under an hour; RPO near zero.
• Multi-site active/active — full capacity in multiple Regions serving traffic simultaneously. Fastest recovery, highest cost.

Choosing for the exam

The question almost always includes a budget or RTO hint. Tight RTO (<1 hour) + cost-conscious → warm standby. 'Can tolerate hours of downtime' → backup & restore or pilot light. 'Zero downtime, cost no object' → active/active.

Remember the supporting services: continuous replication (RDS cross-Region read replicas, Aurora Global Database, S3 CRR, DynamoDB global tables), Route 53 failover routing for traffic shift, and AWS Backup for centralized backup policies.

Common traps

• Multi-AZ is NOT disaster recovery — it protects against AZ loss, not regional loss.
• Cross-Region read replicas alone do not fail over — you need promotion + DNS/traffic shift.
• A snapshot in the same Region does not protect against regional failure.