Question

A media company serves a static single-page application and its API from CloudFront. The origin is an S3 bucket in us-east-1 for static assets, and a Regional Application Load Balancer in us-east-1 for the API. Leadership requires that if the entire us-east-1 origin infrastructure becomes unavailable, viewers automatically continue to be served from a warm standby stack deployed in us-west-2 — without requiring any DNS TTL wait or client-side changes. Which CloudFront configuration meets this requirement with the least additional latency during normal operation?

Accepted Answer

Configure a CloudFront origin group for each origin, with the us-east-1 origin as primary and the us-west-2 origin as secondary, and define failover criteria on specific HTTP status codes.

Answer

Create a Route 53 failover routing policy with health checks in front of the origins and point CloudFront at the Route 53 record name as a single origin.

Answer

Use a Lambda@Edge origin-request function that checks origin health on every request and rewrites the origin domain to us-west-2 when us-east-1 fails.

Answer

Deploy AWS Global Accelerator in front of both Regional stacks and set the Global Accelerator endpoint as the CloudFront origin.