DR Pattern Selector
A small resiliency planner that converts RTO, RPO, and business criticality into an AWS recovery pattern while avoiding unnecessary always-on multi-region spend.
Problem
Disaster recovery decisions are often flattened into generic advice. A useful plan starts with recovery time objective, recovery point objective, business criticality, and the cost of keeping standby capacity alive.
Design
- The demo is protected by Cognito and API Gateway JWT authorization.
- Lambda accepts RTO, RPO, and criticality inputs and maps them to backup and restore, pilot light, or warm standby.
- DynamoDB stores a short-lived plan record so the action is auditable without creating permanent state.
- The response includes the suggested AWS services and the cost/recovery tradeoff.
Cost control
The implementation deliberately does not deploy warm multi-region infrastructure for a portfolio demo. It demonstrates the architecture decision process while keeping the running footprint to one Lambda invocation and one DynamoDB write.
Tradeoff
A real production implementation would validate restore procedures, run game days, monitor replication health, and document failover ownership. The demo focuses on the first architecture conversation: matching recovery objectives to cost.