Q: What are RTO and RPO in AWS cloud architecture? Explain the 4 primary AWS disaster recovery strategies, their cost trade-offs, and how you architect systems to achieve sub-minute recovery.

The foundational disaster recovery metrics:

• RTO (Recovery Time Objective): The maximum acceptable duration of downtime before application service must be restored after a disaster. (How long can the business afford to be offline?).
• RPO (Recovery Point Objective): The maximum acceptable data loss measured in time between the disaster and the latest recoverable backup. (How much data can the business afford to permanently lose?).

Ranging from lowest cost to lowest downtime:

• 1. Backup and Restore (Hours RTO, Hours RPO): Regular snapshots stored in Amazon S3/AWS Backup with Cross-Region Replication. Cheapest strategy, but spinning up compute from scratch in a secondary region takes hours.
• 2. Pilot Light (10s of minutes RTO, Minutes RPO): Critical core data is continuously replicated to the secondary region (e.g., Amazon Aurora read replica). Compute templates (AMIs, Terraform) are ready, but EC2 instances/EKS node groups are scaled to 0 until failover.
• 3. Warm Standby (Minutes RTO, Seconds RPO): A scaled-down, functional duplicate of the system runs continuously in the secondary region. During disaster, Route 53 redirects traffic while Auto Scaling groups rapidly scale compute up to 100%.
• 4. Multi-Region Active-Active (Near-zero RTO, Near-zero RPO): Full production workloads run simultaneously in two or more AWS regions. Route 53 latency/geo-routing distributes live user requests. Amazon Aurora Global Database or DynamoDB Global Tables provide sub-second cross-region replication.

Technologies enabling enterprise DR automation:

• Amazon Route 53 Application Recovery Controller (ARC): Automated health checks, DNS failover, and routing control cells to isolate regional failures.
• Amazon Aurora Global Database: Storage-level cross-region physical replication with typical latency < 1 second and zero performance impact on the primary cluster; promote secondary in < 1 minute.
• DynamoDB Global Tables: Fully managed multi-master active-active database with automatic multi-region replication.
• AWS Backup: Centrally managed cross-region and cross-account immutable vault backups (protected against ransomware).

Failover workflow for Warm Standby or Pilot Light:

• Conduct quarterly 'GameDay' disaster simulations in staging environments to verify that automation works without manual intervention.