CDN Architecture
Serve content from edge nodes geographically close to users, drastically reducing latency and origin load.
★★★★★5/5Network layer — DNS, CDN, load balancers, routing
Interactive visualization
LiveHow it works
A Content Delivery Network (CDN) is a globally distributed network of proxy servers (Points of Presence / PoPs) that cache and serve content closer to end users.
When a user requests an asset, DNS routes them to the nearest PoP. If the PoP has it cached (cache HIT), it's served immediately with sub-millisecond latency. On a cache MISS, the PoP fetches from the origin, caches it, and serves it.
Modern CDNs (Cloudflare, Fastly, CloudFront) go beyond static assets: edge computing (Cloudflare Workers, Lambda@Edge) runs code at the PoP; dynamic acceleration optimises routing for non-cacheable requests; anycast routing distributes load globally.
Why it matters
Network latency is proportional to physical distance. CDNs collapse this distance for users worldwide, improving load times and absorbing traffic spikes that would overwhelm origin servers.
✓ When to use
- →Any public web application with global users
- →Static assets (JS, CSS, images, video)
- →API responses that can be cached even briefly
- →DDoS mitigation via edge absorption
✗ When NOT to use
- →Highly personalised or real-time data that can never be cached
- →Internal applications with users only in one region
Trade-offs
Dramatic latency reduction for global users
Cache invalidation is complex — stale content risks
Absorbs traffic spikes at the edge
Costs scale with bandwidth egress
In production
330+ PoPs serving 20%+ of all web traffic
Open Connect CDN delivers 99%+ of streaming traffic from ISP-embedded appliances
Industry adoption
Related principles
Load Balancing
LiveDistribute incoming traffic across multiple servers to maximise throughput, minimise latency, and prevent overload.
DNS Resolution
LiveTranslate human-readable domain names into IP addresses via a distributed, hierarchical lookup system.
Cloud-Native Architecture
Design applications specifically to exploit cloud capabilities: elasticity, managed services, and pay-per-use scaling.