Delivering high-quality video to consumers is critical for providers of video content, even if it is delivered via the Internet. This is even more true if the provider is a well-known video brand, with viewers used to viewing that content on broadcast or pay-TV networks.
But delivering high quality of experience (QoE) over the Internet is a challenge, with shared bandwidth and unpredictable traffic volumes. The streaming service provider also has no control over the network itself, at least not end-to-end. That's where content delivery networks (CDNs) can help -- they cache content closer to the consumer, and take away some of the uncertainty, distance and resulting delays involved with Internet streaming.
According to Jeff Webb, principal streaming architect at UK broadcaster Sky , "QoE [is] the biggest challenge, and by a very long way. As a broadcaster, it's important that we offer the best QoE and it is consistent across screens and screen sizes."
"One way to ensure QoE is to bring content closer to the customer. It cuts latency and reduces bandwidth requirements," he said.
But what is the best approach to CDNs? Are you better off building your own CDN, or using a commercial one, like an Akamai Technologies Inc. (Nasdaq: AKAM), Level 3 Communications Inc. (NYSE: LVLT) or Limelight Networks Inc. (Nasdaq: LLNW)? Or should you use them all and spread the risk, in a manner of speaking?
Andy Burnett, director of online technology and operations at UK commercial broadcaster ITV plc (London: ITV) advocates using a single CDN. He recommends sitting down with that vendor and working through in great detail what is needed, to ensure you get what you want. This gives the content provider the benefits of high performance without the associated costs and complexity of managing multiple providers.
Burnett did concede there were advantages to a multi-CDN approach, in terms of redundancy and potential efficiency in some situations. But he felt that the trade-off with higher costs made these benefits questionable, and having just one provider was a lot easier.
He had no interest in building a CDN, saying it would take a significant engineering effort to do so and he would rather use that effort to create something that differentiated the service. In fact, he wanted to see the CDN as a commodity, freeing up his development resources to target higher layer features, and thereby differentiating ITV services more effectively.
Burnett's thinking is unusual, with most others in the industry advocating a multi-CDN approach. According to Jon Alexander, senior director of product management at Level 3, multi-CDN deployments are a "best practice for large-scale content delivery."
It's now a requirement for business continuity and disaster recovery, in his opinion. Plus, the ability to split the load across multiple CDN providers helps during periods of unexpected traffic spikes.
"The primary benefit of a multi-CDN strategy is performance-based load-balancing that takes advantage of the strengths of each provider to deliver a better experience than any individual provider could deliver," he said.
But he also agreed to some extent with Burnett, saying that as your CDN count increases, the complexity of troubleshooting, configuration and synchronization involved increases costs. Diminishing returns can set in quickly as the complexity of managing additional CDN providers more than offsets the benefits.
Building your own CDN is also an option, said Jamie Panagos, senior director of video operations at US cable operator Charter Communications Inc. -- in fact, for Charter's needs, he felt it was the only option.
"It's easier to decide to build your own CDN. Scaling a CDN is hard; Akamai, Level 3 all have had problems."
Charter has opted for a very simple CDN, which according to Panagos can scale almost indefinitely for pennies, compared to traditional third-party CDNs. He also noted that the customization work Charter would have had to do for a commercial CDN was almost as much work as building its own.
At Sky, Webb has opted for a mix of third-party CDNs and an in-house CDN, built and managed by Nokia's video division, Velocix . The reason for a multi-CDN strategy is simple; he wanted redundancy and flexibility because the volume of content Sky streams "could break any single CDN."
But why build your own when you are working with multiple CDNs anyway? Webb explained that independent CDNs had their own margins to worry about, and at some point you just couldn't push their pricing any further. With video volumes growing incredibly fast, it made more sense to just build your own, where you don't have to worry about a third-party protecting its margins.
Davide Gandino, OTT & cloud processing & delivery manager, and Webb's compatriot at Sky Italia, also stressed the need for data and flexibility to take advantage of a multi-CDN strategy. Sky Italia has developed a "CDN Selector," which identifies the right CDN to use for every stream, based on a variety of metrics. This allows Gandino to track various key performance indicators (KPIs) such as the buffer ratio, average bit rate, 404 or other HTTP errors.
"We know if there is an issue on a particular ISP ... or we know that there is an issue on a specific CDN ... we use five or six CDNs, so we decide to use the best one for every specific client, every specific user," he said.
Both Webb and Alexander also stressed that different CDNs performed differently on different ISPs. Geographic coverage and traffic congestion are most certainly important issues, but the proximity of caches to the ISP network, peering relationships etc., could also make a difference.
According to Alexander, "It is important to realize that each CDN has strengths and weaknesses in different geographies, networks and devices and therefore the multi-CDN load-balancing system used should have the granularity to allocate traffic between providers based on the intersection of these three factors."
As chief strategist at Akamai's mobile business unit, Michael Archer advocates an additional step in taking network unpredictability out of the equation. He recommends using a pre-loading approach, aimed primarily at mobile devices. The content is downloaded based on a set of rules set by the content provider and the consumer (For example: only when plugged in, only over WiFi, only between 3 a.m. and 6 a.m., etc.) and is based on the consumer's usage profile and content preferences, so only relevant content is delivered. This eliminates any QoE issues stemming from network connectivity, and content can even be viewed when the consumer is not connected.
So what is the best CDN strategy? Unfortunately it doesn't seem like there is a clear answer. Ian Parr, head of design at BT, said that it was difficult to fix on a single model, as the market kept changing that equation. The optimal approach would vary based on the provider's specific needs and the technology and pricing options available at the time. He pointed out that BT now has a variety of approaches on its network. It has Netflix and Google caches embedded into its network, and also uses Akamai and Limelight to distribute its sports service online. But this is a change of strategy for the operator, having built its own CDN using Cisco Systems Inc. (Nasdaq: CSCO)technology some years ago.
Echoing Webb's comments about QoE, Parr stressed it was critical that providers get it right. For BT, 70% of traffic on the network is already video, and now, as a sports broadcaster, high QoE video delivery had become even more important.
Burnett added that some ITV shows already generate more traffic online than on the linear channel, which made him think that online video was going to grow to a factor the industry can't even predict today.
So while determining the right approach for your organization may be dependent on individual requirements, developing that CDN strategy is critical for any multiscreen video provider.
As they never quite say on Game of Thrones -- "Video is coming."
Note: Comments in this article have been taken from prior interviews and statements at industry events by the executives mentioned..
— Aditya Kishore, Practice Leader, Video Transformation, Telco Transformation