CDNs have always been a necessity in the world of streaming video. Although some streaming providers, such as Netflix, Apple, and Amazon, have continued to build their own delivery infrastructures, the majority of streaming providers utilize a combination of CDNs to ensure that their content is delivered in the best possible way to viewers. That’s not to say that these other providers, like Disney+ and Peacock, don't use their own CDNs as part of the delivery mix (or can't afford to build their own). Rather, it’s a strategic decision. Netflix deployed its own CDN technologies into carrier networks as streaming began to surge, which gave it the opportunity to fine-tune and tweak them as streaming grew. But providers like Disney+, Peacock, and HBO Max have dropped into a cresting wave. Deploying their own CDN technologies in lieu of those of commercial providers would be a Herculean task—and an awfully expensive one to get enough infrastructure set to handle hockey-stick subscriber and viewer growth.
It’s probably safe to say that without CDN technologies (like reverse proxies and cache hierarchies), streaming would be much different today. The large commercial CDNs have massive networks, sometimes extending well inside the ISP, and ensure that popular content is cached much closer to the viewer than it otherwise would be. Without these networks and those deployed by companies like Netflix, streaming scale, quality, and availability might have been significantly hampered. Of course, the move from Real-Time Messaging Protocol (RTMP) to HTTP helped as well, by enabling existing web delivery servers, rather than proprietary media servers, to take part in delivering streaming content.
But streaming is still only a fraction of total video watched, even post-COVID lockdown. What happens when it's the de facto method for delivering video content? Will the existing CDN approaches be enough to handle the next phase of streaming growth?
Of course, there are a variety of ways to architect a CDN. Some, like Akamai’s, utilize algorithms to route around internet congestion. Others, like Limelight Networks', use a lease-wavelength private network. And other companies approach content delivery in their own unique way as well: Fastly, Amazon, Verizon, CenturyLink, StackPath, etc. In addition, we are starting to see combinations of approaches. Look at CenturyLink's recent acquisition of Streamroot, a peer-to-peer CDN. Although peer-to-peer was panned in the last decade because of some insidious software used to seed peers and its difficulty in securing content, it evolved as a critical delivery component for live-streaming traffic.
CDN technologies are evolving, though. Companies like Eluvio, with its "content fabric," are attempting to redefine the process through which content is made available to end users. It's not just about more infrastructure; it's about using infrastructure differently. But it's also not the only way that the CDN is evolving to meet future streaming demands. As the edge continues to grow in importance to streaming, CDNs will naturally make use of resources that continue to extend closer to the end user in the cloud, mist, and fog. It won’t just be about cross-connects and peering fabrics in the future of streaming delivery.
It will be about massive, parallel, distributed delivery that includes not only commodity and containerized systems, but most likely everything from network-attached storage in the home to the devices we have in our hands. To meet the scalability, resiliency, and quality expectations viewers have for streaming (Hint: Consumers don’t care if a video is broadcast or streamed, just that it’s flawless, uninterrupted by quality issues like buffering, and always available), CDNs will have to break out of the traditional cache/parent-cache/origin architectural model to embrace something that provides delivery on a smaller and more efficient scale.
The CDN is the backbone of streaming, the "coaxial" of the internet video experience. What the future of the CDN will be is anyone's guess, but one thing is for certain: Its evolution will be something to watch.
A blockchain-based, ledger-and-wallet approach promises a digital rights management that's just as effective as anything we have now, with the benefit of being more consumer-friendly.
17 Jun 2022
The Asia market is different from Europe and the Americas in many ways, especially in terms of user behaviors and the internet environment. CDN providers need to deal with many nuances to succeed and capitalize on the market opportunity.
06 Nov 2020
With a clear understanding of traffic, key performance indicators, and what you need to store on an origin server, a multi-CDN strategy can improve a better experience.
19 Feb 2020
Delivering content at scale requires a number of precise, discrete, yet interconnected steps, starting with an acute awareness of server loads.
23 Jan 2020