Using multiple CDNs is a way of life in today’s streaming world. From scalability to resiliency to ensuring the best quality of experience, many streaming operators have built powerful business logic to not only help them evaluate CDNs in real-time but switch between them on a user-by-user basis.
Unfortunately, none of that solves the underlying issue: using multiple CDNs significantly complicates operating a streaming service. Overcoming the challenges of multi-CDN strategy requires more advanced business logic and different processes. It also necessitates a way to deal with the impact that using multiple CDNs has on your operations team’s ability to troubleshoot issues quickly and provide the best quality of experience.
Multi-CDN strategy challenges: Dealing with one CDN is hard enough…but 6?
You already know what it’s like to work with one CDN when troubleshooting delivery issues. It requires coordination with the CDN, ISP, and even other vendors just to figure out what is actually causing the problem. Now multiply that by 6. It becomes almost untenable to troubleshoot when content is being delivered by six different networks.
Moreover, it’s not that the issues are just harder to find, it’s that you may have different issues simultaneously across multiple networks but only one operations team to sort through them. Which issue is fixed first? Which users get to experience a degraded viewing experience the longest? Addressing those issues is complicated by the fact that every CDN provider has its own set of tools for configurations, purging, traffic analysis, regexes, etc. When operations teams have to jump from dashboard to dashboard and implement fixes in different ways for different CDNs, it slows issue resolution way down.
So how do you solve operationalising your multi-CDN strategy in a way that doesn’t slow everything down? It’s not a matter of getting access to more data or more dashboards. It’s about your operations team becoming smarter about how that data is collected and visualised. Otherwise, finding the root cause and troubleshooting issues takes more time, which affects your QoS, your capability to meet SLAs, and ultimately, your revenues.
How to operationalise your multi-CDN strategy
Two things are needed to operationalise supporting multi-CDN delivery: consolidation and visualisation.
With regard to consolidation, operations teams need to pull all of the CDN data into a standardised framework. Ops engineers can’t look at data for one CDN differently than another. This can further complicate the troubleshooting process. If the same issue exists on CDN 1 and CDN 2 but the data used to diagnose it is different, it lengthens both MTTD and MTTR. Yet if all the data is brought into a single location, there’s an opportunity to set up business rules to normalise the data against a standard data dictionary the ops team might employ.
It’s not just CDN data, though. A good multi-CDN strategy also needs you to bring in the data from other components in the tech stack. Having everything in a single data pool is critical to creating the kind of connections that will help identify issues before they happen. Yet that’s impossible if your ops engineers are bouncing from dashboard to dashboard, trying to make connections themselves.
Once all of the data is in a single place, the next step is to ensure proper visualisation. Data has to be displayed in a way that makes issue investigation the same regardless of the provider, regardless of the original source of the data. This allows the operations team to develop and apply a single process to issue analysis and resolution which can radically speed up MTTD and MTTR.
The key multi-CDN solutions to get you from where you are now to where you need to be
Your multi-CDN solution needs three things to get to that place where you can consolidate your CDN data into a single visualisation tool: agents, a monitoring harness, and a visualisation tool.
Agents
Firstly, you’ll need to deploy agents to monitor your CDNs. It’s not enough just to get log data or access data via APIs. You’ll need to coordinate that with data taken from agents monitoring your streams independently. This kind of synthetic monitoring will provide an objective view against the data provided by each CDN.
Monitoring harness
You’ll need a programmatic framework that enables you to easily plug in new data sources, like an additional CDN, without having to re-develop your dashboard. This ensures your operations team doesn’t have to continually learn new visualisation tools and can apply established processes for analysing data and troubleshooting issues regardless of the data.
👉 Read more on this in our Monitoring Harness whitepaper.
Visualisation tool
Last, but certainly not least, you’ll need a visualisation tool that makes diagnosing errors across CDNs (and the workflow) much easier. Imagine a red, yellow, green system that clearly identifies trouble spots regardless of their location in the workflow: upstream from the CDN or downstream at the player.
The real benefits of an effective multi-CDN strategy and delivery
Sure, with that consolidation and visualisation, you’ll be able to troubleshoot potential service and quality issues much quicker and more efficiently. Yet there are other benefits to employing this kind of strategy for operationalising multi-CDN delivery.
Uninterrupted streaming experience
The goal is a reliable, scalable streaming service that “just works.” You know you need multi-CDNs, but they don’t have to interrupt your operational team’s ability to ensure great service.
Reduction or even prevention of downtime
The biggest contributor to service interruption or downtime is the inability to quickly identify and resolve issues. With a simple visualisation tool that draws in data from across the workflow, including external sources like all those CDNs you are using, you can address issues before they become service interruptions. Just imagine: proactive, not reactive.
Improved performance
By identifying errors and issues before they happen, the performance of your service will improve. Collecting, normalising, and displaying all of your streaming workflow data in an easy-to-use visualisation tool offers the opportunity to not only solve problems but optimise components when its data falls outside of acceptable ranges.
It’s important to note, though, that this approach is more than just about CDN data and operationalising multi-CDN support. It’s also about improving collaboration with vendors. By providing them access to the same dashboard, so that they can see a yellow or red element that represents their product, you can work with them to more quickly and efficiently identify the problem. This is the other key benefit of this approach: it’s a way to bring all of your workflow data together. From CDNs to encoders to origins to players, you can visualise all of it together so that your operations teams can achieve true observability–seeing problems before they contribute to subscriber dissatisfaction or, worse, churn.
Tame your multi-headed monitoring monster
A monitoring harness. Synthetic monitoring agents. A unified visualisation tool. With the proper monitoring infrastructure in place, your multi-CDN strategy doesn’t have to create multi-problems in your ability to provide a reliable, scalable, high-quality viewer experience. Only putting those monitoring pieces in place yourself might prove just as problematic as managing multiple CDNs without it.
Touchstream can help you operationalise the monitoring of your CDNs–and your entire streaming workflow–with a platform you can use today but will grow with you tomorrow.
Ready to build a multi-CDN strategy that works? Book a demo now.
