Content Distribution Networks (CDNs) and cloud providers are a crucial part of your content delivery strategy. To ensure a positive end-user experience and low latency data delivery for global audiences, an optimised CDN is a must-have. There are an array of providers who offer quality solutions, but those from Cloudflare, Akami, AWS & Google Cloud are the more prominent.
In this article, we illustrate how to optimise your distribution strategy. First, we briefly cover what a CDN does. Next, we explain what metrics to monitor to maintain an optimised CDN. Finally, we explain the crucial step that is too often ignored by organisations in their content distribution strategy. Ignoring this step can compromise a lot of the hard work done to optimise your CDN or content distribution strategy.
The chief reason for using a CDN is to enhance your user's experience in the matter of speed. People expect their films, games, financial data or communications to be available on-demand, at all times, in any location and on every device we have. It is impossible to fulfil these demands without an optimised CDN (be it a self-deployment or a commercial offering).
CDNs can also aid in greater service resilience by reducing the load on isolated components. In the event of traffic surges, CDNs help to distribute bandwidth across multiple servers, instead of allowing a single server cluster to deal with all traffic alone.
The benefits of an optimised CDN and stack distribution are therefore significant.
With video streaming & OTT broadcasting, the aim is to have a consistent and near real-time stream so that footage has little to no buffering and delays. Given that video is exceptionally data-intensive and requires extensive bandwidth, an optimised CDN is even more critical in providing a satisfactory user experience.
Given the importance of CDNs, they must be optimised to ensure you can deliver consistent content and deal with any level of requests at any time. Here is a list of metrics and data to monitor to ensure your CDN is consistently performing to your needs:
We have covered the importance of CDNs in delivering your content or video streams with minimal latency to your global audiences. We have also explained how to optimise your CDN best. However, there is one crucial step rarely mentioned in the case of content distribution via a CDN or your cloud provider: how your data gets to them in the first place.
By now, you have likely invested a lot of time and resource in finding efficiencies with a cloud provider or CDN operator to make it suitable for your data transfer. Yet, why go to all this trouble if your content is travelling to it on a b-road? All of your hard work optimising your CDN can be easily undone by having poorly optimised entry onto other platforms.
Let's look at this from the perspective of live sports broadcasting. Before we even consider global distribution, many simultaneous things need to happen. For example, footage from the stadium, commentary from an off-site production centre, and visual overlays from another location must all come together first, to form a finished product. These need to arrive at your CDN ingress in the most efficient way possible. If your connection to your cloud provider/CDN is inadequate, your hard grafted CDN optimisation suddenly becomes compromised. Sure, it can optimise the distribution of your finished content, but only once it has received it. If this is taking too long, there is little point in having that finely tuned CDN.
How you connect to your CDN is almost as crucial as having a CDN itself. They go hand in hand. The journey your data travels before it reaches the CDN is a fundamental piece in your content distribution strategy, too often overlooked.
1. Connecting via the public internet
We shall only briefly cover this as a viable option because it is futile if you need to transfer high capacity, data-rich content. Using the public internet puts you at the mercy of global ISP routing and peering. It does provide resiliency, but it is inherently unpredictable. It would be best if you had the assurance of how and where your data is travelling, and how long it takes to do so.
Something essential to consider on top of this and often neglected is the impact on poor intra-region connectivity. Routing your data from one region to another can be similarly troublesome as over the public internet. Your cloud provider has to juggle your traffic amongst many other customer requests. The result? Minimal transparency in geographic routes, and once again, loss of control of your connectivity.
Using the internet (over IP-SEC or otherwise) for transport can be a low-cost option for accessing public cloud services. However, for delivering time-sensitive and high capacity data, it is challenging in terms of speed, latency, security, and control. Given the importance of end-user experience, you want maximum control over how it arrives there.
So, the only way to guarantee this performance is to establish a private connection.
2. Direct connection to the cloud
The "big three" cloud service providers, Amazon Web Services, Microsoft Azure, and Google offer ways to resiliently connect to their public clouds by establishing dedicated ports. The AWS version of this is Direct Connect, Microsoft Azure calls it ExpressRoute, and Cloud Interconnect is Google's.
A specialist network provider can design dedicated routes between your office or Data centre to ensure your transmission is both performant and resilient.
Dedicated routes can also reduce costs as metered egress off of your cloud provider can be at a lower rate. It also increases bandwidth throughput and provides a more consistent network experience than Internet-based connections.
There are variations between each platform regarding redundancy and access methods, but they all allow your cloud resources to be routable directly from within your enterprise network. You are in full knowledge of what hops your data make, and you have a clear expectation of how long they should take.
Working with a specialist network provider also brings you additional benefits on top of those mentioned above. For example, at 01T, every route is offered fully managed and monitored for you: tickets proactively raised, thresholds set with instant notifications of any breach, and a consistently reliable and helpful project manager to talk to directly, any time. Every single network service is 100% (1:1) uncontended, too.
You can chat with one of our technical architects if you have any questions on content delivery, or if you would like to find out more about how a specialist network provider can help you.
We look forward to hearing from you.