If you already have a multi-CDN strategy in place, it’s time to think about investing in a smart multi-CDN switching solution that allows you to make the most of your multi-CDN configuration at all times.
Most multi-CDN switching tools available today are not built with video streaming in mind, falling short of the needs and challenges streaming providers face in their day-to-day.
Selecting the best CDN from your portfolio for each video chunk, user, device, and region is a complex process that requires the speed and automation only an intelligent and flexible system can provide.
Here are the 4 key steps toward implementing a smart multi-CDN switching tool for video streaming.
1. Accurately identify traffic and end-user information
The most common type of CDN load balancing solution used today are systems based on the Domain Name System (DNS). Although easy to implement, DNS-based technologies have limitations when it comes to accurately identifying end-user traffic information.
On the other hand, a next-generation multi-CDN switching tool that feeds directly from the video player is able to accurately identify all traffic and end-user information. Such a specialized tool can track traffic attributes ranging from the type of user (e.g. premium or not premium) and region to the type of content or Internet service provider (ISP).
Good traffic identification allows you to establish CDN switching rules that take into account the full picture of your platform’s traffic. You need exact traffic data to make the right switching decisions; the more information you have, the better.
With proper end-user identification, a smart multi-CDN switching tool can then optimize content delivery for each network segment. The right CDN can be selected for each chunk of video, user type, device, region, and ISP
2. Leverage real-time quality of experience insights
As opposed to most CDN switching solutions in the market, a multi-CDN balancing solution made specifically for video streaming like that of NPAW is able to track quality of experience in real time across CDNs.
Without real-time quality data, a balancing system cannot make smart decisions. CDN switching is limited to a more or less sophisticated round-robin, in which the next CDN in line is selected when the current one fails. This gives a false sense of security, as, without accurate quality of experience data, there is a chance the wrong CDN will end up being selected.
On the other hand, a smart multi-CDN balancing system based on real-time quality of experience insights allows streaming providers to switch between CDNs based on actual facts rather than educated guesses.
For example, NPAW’s CDN Balancer leverages the extensive video insights tracking capabilities of the NPAW Suite of video analytics solutions to dissect traffic into thousands of network regions, segmenting it in real-time based on the quality of experience for any combination of region, city, ISP, protocol, and origin.
3. Apply complex QoE-based business rules with laser precision
Segmentation, flexibility, and precision go hand-in-hand when it comes to smart multi-CDN switching. Being able to identify different network segments and needs isn’t worth much if you can’t easily and precisely implement rules that take them all into account while protecting your business interests.
Let’s say you want to use a combination of CDNs in one region, and a different one in another region. For both regions, you want to select the best CDN in terms of quality, but the cheapest in case various CDN present similar quality levels — all while making sure that you cover all the traffic commitments with each CDN provider and avoid any overage bills.
But precision is also important. It is not necessary to stop using a CDN in an entire region if that CDN has a problem in only one city or an ISP within that city.
It is critical for streaming services to have a solution that is capable of accommodating these and many other needs in real time through a simple interface. This requires a CDN balancing system that can analyze a series of KPIs obtained simultaneously from a myriad of different data points — at NPAW we have scenarios where we calculate KPIs for more than 500,000 different combinations (CDN/region/city/ISP/protocol/origin) every minute.
4. Avoid a single point of failure with multiple balancing methods
An adequate multi-CDN switching system must allow for different CDN switching options to ensure the correct distribution of content in case of catastrophic failure. After all, one of the main reasons for implementing a multi-CDN strategy is to provide greater protection against failures, so the balancing system cannot in any case introduce a single point of failure in the chain.
As a minimum requirement, a smart multi-CDN balancer should deliver three different types of switching.
First, it should offer a DNS-based CDN switching option. Easy to implement, DNS-based solutions are the most ubiquitous in the market and allow for the balancing of both video and static page loads. However, CDN performance and quality of experience cannot be tracked in real time. Plus, they represent a single point of failure: in the case of DNS failure, the entire balancing system goes down.
These limitations present the need for a second balancing method based on an application programming interface (API). This switching method permits a much more granular selection of CDNs by having a better knowledge of the traffic to be balanced, as well as by allowing total control of the system’s operational aspects.
Finally, a mid-stream CDN switching solution like NPAW’s CDN Active Switching lets you apply complex business rules while also being able to automatically detect CDN failures before they happen and switch amongst the originally designated CDNs for seamless, transparent switching.
