Your e-commerce backend on Cloud SQL is overloaded, causing API latency and timeouts. You control the API. Which solution optimizes API performance and user experience?

This solution directly addresses the core problems of API latency and database overload. Apigee provides an API gateway for managing and securing API traffic. Memorystore (Redis) is an in-memory data store ideal for caching "hot" (frequently accessed) data, significantly reducing the load on Cloud SQL and improving response times. Exponential backoff for retries in the application prevents a thundering herd problem, where many failed requests retry simultaneously, further overwhelming the backend. The other options are less effective: Enforcing rate limits and queuing requests with Pub/Sub (option 2) can protect the database but doesn't inherently reduce latency for valid requests or improve performance. Cloud CDN (option 3) caches static content, not dynamic API responses from a backend database, so it wouldn't help with the database overload. Increasing DB instance memory (option 4) might offer a temporary boost but doesn't fundamentally solve the issue of inefficient data access or high request volume, and custom rate limiting adds complexity without addressing the root cause of database strain.