CDN Architecture and Points of Presence
What You'll Learn
You'll understand how CDNs physically distribute content around the globe using edge locations and Points of Presence (PoPs).
The Physical Infrastructure Behind CDNs
Think of a CDN as a global network of strategically placed warehouses. Instead of shipping products from one central warehouse (your origin server) to customers worldwide—which would be slow and expensive—you stock smaller warehouses (edge locations) closer to customers in major cities.
Points of Presence (PoPs)
A Point of Presence (PoP) is a physical data center location where CDN infrastructure exists. Each PoP contains multiple edge servers that cache and serve content. Major CDNs operate hundreds of PoPs worldwide—in cities like Tokyo, London, São Paulo, and Sydney.
Edge Locations vs PoPs
- PoP: The physical facility housing infrastructure
- Edge Location: The servers within a PoP that actually cache and serve content
Some providers use these terms interchangeably, but technically a single PoP may contain multiple edge locations or server clusters.
How Content Flows Through the Architecture
- Origin Server: Your source of truth (your application servers or storage)
- Regional Edge Caches: Mid-tier caches serving multiple edge locations (not all CDNs have this tier)
- Edge Locations: Front-line servers closest to users
- End Users: Request content from their nearest edge location
When you request a video, the CDN routes you to the nearest PoP. If that edge location has the content cached, it serves it instantly. If not, it fetches from the origin (or a regional cache), stores a copy, and delivers it to you—and to future visitors from that region.
Key Takeaway: CDN architecture relies on a globally distributed network of PoPs containing edge servers that cache content close to users, reducing latency and origin server load by serving cached copies from the nearest location.