A content delivery network (CDN) increases network performance and uptime by caching frequently-accessed content on local servers that are positioned at the Internet’s edge.
CDNs are a vital backbone for much of the data that traverses the Internet. In fact, the majority of web traffic today is served through a CDN, including huge portions of web traffic commanded by companies such as Facebook, Netflix, and Amazon.
In addition to performance advantages, CDNs reduce bandwidth demands on a website’s origin server, which helps reduce costs, cut down on service interruptions, and provide better defense against threats like DDoS attacks.
As the cloud continues to become an increasingly dominant model for data delivery, the demand for CDNs will continue to accelerate. The simple distance between end-users and cloud servers introduces latency that can have a detrimental effect on a broad range of use cases, from real-time B2B services to gaming and streaming.
Because of this, bringing content closer to the end-user is vital, which is what CDNs have achieved since the first one was created in the late nineties.
Internet Exchange Points (IXPs) are a set of global connection hubs for web traffic originating from different internet service providers. A CDN maintains storage locations known as Points of Presence (PoPs) in close proximity to these vital hubs.
By putting key content near region-specific IXPs with extensive global network integration, content is delivered more efficiently than it is from an origin server located in a distant datacenter. In addition to sheer proximity, CDNs provide load balancing, solid state storage, and file compression.
Step by step, here is how a CDN works:
Imagine that a US retailer is looking to expand into the East Asian market. As a user in Asia browses their e-commerce storefront to check product availability, real-time data needs to be pulled from a diverse set of enterprise systems. Required data may need to be integrated with systems covering pricing, marketing, sales, logistics/warehousing, and customer service databases and apps.
In a traditional architecture, this data would need to travel from the company’s originating server to the relevant regional IXP, across the Pacific to another IXP for the relevant region in Asia, and then finally on to the end-user’s ISP and device. Even making optimistic assumptions about hardware uptime, bandwidth, and speed at each stop along the way, this sort of circuitous route is certain to introduce increased latency.
By deploying a CDN, this company could transform a clunky digital experience for Asian users into a highly responsive one capable of supporting dynamic marketing and support strategies.
This is a hypothetical example, but there are thousands of real world examples, including those documented in the following StackPath case studies:
A CDN has a close functional similarity to an edge network as each describe architectures that put key technology resources closer to the end user. The main difference is that a CDN focuses more on delivering data while edge networks are capable of projecting broader computing services to the edge.
In cases like StackPath, that means a CDN is a fully integrated part of a broader edge network implementation. With this overlap mind, we can view CDNs as a subset of edge networking, though CDNs are a resource that many enterprises take advantage of without a broader edge-facing strategy.