The History of the CDN – and Why It Matters
By Ted Middleton, Chief Product Officer
When content delivery networks (CDNs) started back in the late ‘90s, the internet was a very, very different place in ways too numerous to count: typical user bandwidth, cost-of-network bandwidth, supported video formats, browser capabilities and implementation variances, web server architecture, security, the list goes on. Entire industries have “gone online” in that time span. The internet evolved from early geek-centric tools like Telnet and Gopher, to Compuserve and AOL and email, which enabled the masses to participate, to the web browser and marketing presence websites, to commerce, to media, and on to cloud and IoT, app stores and voice assistants. The internet and web have moved from being “simply” a communication and information sharing tool, to a fundamental fabric upon which almost every aspect of modern society and huge portions of the global economy depend.
Similarly, CDNs started out simply; largely as a bunch of servers around the world, connected up to the internet, with smart software routing requests to the appropriate server, which cached and served up content as needed. Media has long been a staple of CDN, and recent shifts in standards, consumption behaviors and consumer expectations have reached a tipping point where CDNs are now fundamental to most of our media consumption and commerce. Next came security services and processing capabilities like dynamic page rendering and image optimization as consumption patterns shifted to mobile and other connected devices.
Through all of these innovations and enhancements, the simple CDN evolved into a rich portfolio of services, doing a lot more than caching content at the edge of the network, and always evolving to meet their core tenets of achieving maximum performance, scale and reliability. CDNs help make the internet what it is today. They are fundamental to the way the internet works and the way that content providers think about (or don’t have to think about) building and deploying their applications. Less apparent to many is the depth of technology and software innovation required to build, operate and maintain a CDN at scale. Constant innovation is a must if a CDN is to grow and support the evolution of content distribution and consumption.
CDNs have gone from hundreds of megabits per second to tens of terabits per second, in terms of capacity, and from millions to billions to trillions of transactions and requests per day. To maintain performance and reliability in a massively distributed environment, all while addressing various scaling and technology barriers, the CDN must keep deploying new code and continually re-engineer and re-architect components of its infrastructure.
The key point in reviewing this history is this: CDN in concept seems easy and obvious now. CDN in practice, at scale is hard. Experience and pedigree counts. Deep expertise and software innovation is a prerequisite for success. Full-stack development and control over every detail from kernel to application and from server port to core router is key to performance, reliability and scale.
Our history and experience prepare us well to adapt to the continually changing nature of the internet, and we never stop innovating. Our CDN customers are always building something new, and by working in partnership with them, we face both the opportunity and the challenge of inventing exciting, new technology and services. The CDN has come a long way in the past two decades, and we’re taking it further to give content owners and application developers the reliability and performance necessary to offer even richer and more engaging experiences – the kind that enable millions of innovators and billions of users worldwide to make the most of the fabric of the modern world we call the internet.