Disaster Recovery and Failover Design
Implement a working warm-standby DR pattern across two regions: cross-region backups with AWS Backup, a standby RDS read replica, and Route 53 health-check-based failover routing — then simulate an outage and measure your actual RTO.
Prerequisites: An AWS account, AWS CLI v2 configured, a registered domain (or a Route 53 public hosted zone you control) for the failover routing exercise, and the RDS instance from the RDS and Database Backups project (or a fresh one).
AWS describes four DR strategies on a cost/recovery-time spectrum: Backup and Restore (cheapest, hours of RTO), Pilot Light (minimal standby infrastructure, tens of minutes), Warm Standby (a scaled-down but fully functional copy running in the second region, single-digit minutes), and Multi-Site Active/Active (both regions serving live traffic, near-zero RTO, most expensive and most complex). Which one is right depends entirely on your RTO (Recovery Time Objective — how long can you be down) and RPO (Recovery Point Objective — how much data can you afford to lose) requirements, which are business decisions, not technical ones.
This project implements Warm Standby, the pattern most real teams land on because it's a meaningful step up from Backup and Restore without the operational cost of full active-active. You'll build cross-region backup coverage, a live (if scaled-down) standby database, and DNS-level failover — then actually break the primary and time how long recovery takes, rather than trusting the theoretical numbers.
Test what you just learned
Self-testing is one of the best ways to retain new skills. Unlock project quizzes to check your understanding.
Log in to unlock0 / 6 complete