Question

A media company runs a mobile news application backed by a monolithic REST API on EC2 behind an Application Load Balancer. The API aggregates data from three microservices, and mobile clients frequently over-fetch large JSON payloads on cellular networks, driving high data transfer costs and poor perceived performance. Product analytics show that different app screens need different subsets of the same underlying data, and the team also wants to push live breaking-news updates to active users without polling. The company wants to improve performance and reduce redundant data transfer while minimizing custom server infrastructure to maintain. Which refactoring approach best meets these goals?

Accepted Answer

Introduce an AWS AppSync GraphQL API in front of the microservices, letting clients request only the fields they need per screen and using AppSync subscriptions over WebSockets for real-time breaking-news push.

Answer

Add an Amazon CloudFront distribution in front of the ALB with a long cache TTL and enable gzip compression to reduce payload size for mobile clients.

Answer

Replace the REST API with Amazon API Gateway REST endpoints, one per app screen, and use Amazon SNS mobile push notifications for breaking-news alerts.

Answer

Keep the REST API but add an Amazon ElastiCache for Redis layer to cache aggregated responses and reduce backend load, then poll the API every 30 seconds from clients for updates.