AWS Certified Solutions Architect – Professional (SAP-C02) — Question 200

A company runs a web application on AWS. The web application delivers static content from an Amazon S3 bucket that is behind an Amazon CloudFront distribution. The application serves dynamic content by using an Application Load Balancer (ALB) that distributes requests to a fleet of Amazon EC2 instances in Auto Scaling groups. The application uses a domain name setup in Amazon Route 53.

Some users reported occasional issues when the users attempted to access the website during peak hours. An operations team found that the ALB sometimes returned HTTP 503 Service Unavailable errors. The company wants to display a custom error message page when these errors occur. The page should be displayed immediately for this error code.

Which solution will meet these requirements with the LEAST operational overhead?

Answer options

• A. Set up a Route 53 failover routing policy. Configure a health check to determine the status of the ALB endpoint and to fail over to the failover S3 bucket endpoint.
• B. Create a second CloudFront distribution and an S3 static website to host the custom error page. Set up a Route 53 failover routing policy. Use an active-passive configuration between the two distributions.
• C. Create a CloudFront origin group that has two origins. Set the ALB endpoint as the primary origin. For the secondary origin, set an S3 bucket that is configured to host a static website Set up origin failover for the CloudFront distribution. Update the S3 static website to incorporate the custom error page.
• D. Create a CloudFront function that validates each HTTP response code that the ALB returns. Create an S3 static website in an S3 bucket. Upload the custom error page to the S3 bucket as a failover. Update the function to read the S3 bucket and to serve the error page to the end users.

Correct answer: C

Explanation

Option C is correct because it allows for origin failover, ensuring that if the ALB is unavailable, the custom error page hosted on the S3 bucket can be served immediately. Option A introduces unnecessary complexity with health checks and does not provide immediate error page display. Option B involves setting up a second CloudFront distribution, which adds operational overhead. Option D requires creating and managing a CloudFront function, which is also more complex than simply using origin failover.