Raj S

HLD Interview Notes: Day 17/30 Topic: CDN: Content Delivery Networks and Edge Caching Imagine millions of users accessing images, videos, CSS, and JavaScript files. If every request reaches your origin server, latency increases and the server can become overloaded. How do we deliver content faster? That's where a CDN comes in. What Is a CDN? A Content Delivery Network (CDN) is a geographically distributed network of servers that delivers content from locations closer to users. It helps improve: • Content delivery speed • Global application performance • Availability • Origin server scalability • Bandwidth efficiency How Does a CDN Work? Without a CDN: User → Origin Server → User Every request travels to the origin server, even when users are geographically far away. With a CDN: User → Nearby CDN Edge → User The CDN checks whether the requested content is cached. • Cache Hit: Returns content directly from the edge server. • Cache Miss: Fetches content from the origin server and may cache it for future requests. What Is Edge Caching? Edge caching stores frequently requested content on CDN servers closer to users. Example: An e-commerce website serves the same product image to thousands of users. Instead of repeatedly fetching it from the origin server, the CDN stores a copy at an edge location. Benefits: • Lower latency • Reduced origin traffic • Faster content delivery • Better handling of traffic spikes What Is Cache TTL? TTL stands for Time To Live. It determines how long cached content remains valid before it needs to be refreshed or revalidated. Example: A product image is cached for 24 hours. After the TTL expires, the CDN can fetch a fresh version or revalidate the existing copy. Common techniques: • Cache-Control headers • ETag validation • Versioned URLs • Cache invalidation What Is Cache Invalidation? Cache invalidation removes or marks cached content as stale when the original content changes. Example: A seller updates a product image, but the CDN still stores the old version. Common solutions: • Cache Purging: Remove cached content. • Versioned URLs: Use a new URL for updated content. • TTL Expiration: Let cached content expire. • Revalidation: Check content freshness using ETags. CDN vs Origin Server Origin Server: Stores the original content and handles requests when the CDN cannot serve a valid cached response. CDN Edge Server: Stores cached copies closer to users and handles repeated requests. Remember: The origin is the source of truth. The CDN improves content delivery. What Content Should Be Cached? Static Content: • Images • CSS and JavaScript • Fonts • Videos • Documents Dynamic Content: Personalized data, account information, and frequently changing responses require careful caching rules. Incorrect caching can expose private data or serve outdated information. Where Is It Used? • E-commerce platforms • Video streaming applications • Social media platforms • Gaming platforms • News websites • Software download platforms Popular providers: Cloudflare, Amazon CloudFront, and Akamai. Common Interview Mistakes • Confusing CDN with a load balancer. • Assuming every request is a cache hit. • Ignoring cache invalidation. • Choosing an inappropriate TTL. • Caching sensitive data incorrectly. • Forgetting origin fallback. • Assuming a CDN automatically improves database performance. Interview Questions What is a CDN? How does a CDN reduce latency? What is an edge server? Cache Hit vs Cache Miss? What is Edge Caching? What is TTL? How does cache invalidation work? CDN vs Load Balancer? How are cache misses handled? How would you design a global image delivery system? FAANG Takeaway A CDN improves application performance by distributing cached content across geographically distributed edge servers. The key decisions are: • What should be cached? • How long should it be cached? • How should stale content be refreshed? Remember: CDN brings content closer to users. Edge caching reduces repeated requests to the origin server. Next: Day 18/30 Rate Limiting: Controlling Traffic in Distributed Systems