Exploring the Edge in the Streaming Video Workflow

The edge of the network, that fuzzy zone between the viewer and the core of the Internet, has long attracted the attention of innovative engineers and more than one engineering effort has sprung to life, had its moment of buzz, and withered on the vine unadopted. Tapping the potential of the edge has proven elusive. Yet here we are in 2025 and the SVTA has launched a new Edge Working Group. What’s different in 2025 that makes the SVTA believe the time is right for such an effort to finally succeed?

We’re at a beginning at a new phase of streaming – delivery of live sports, in high-quality and low-latency – and this is driving a reworking of streaming architectures; VOD and AVOD streaming architectures can be made to work for small events, especially those that can tolerate higher latency, but once viewership levels push events into medium, large, and especially super-sized levels, even the best VOD/AVOD architectures can become heavily stressed and even beyond the breaking point; streaming video engineers are ready to re-architect for scalable live.

But the shifting economics of streaming delivery simply don’t support the approach of throwing more and more money at shoring up VOD/AVOD infrastructure to meet the scaling needs of ever-growing live event viewership. Business recognizes that they need a better approach, one that scales more cost efficiently, and to continue to keep that cost focus as live viewership grows, with ever bigger events to come. In short, streaming businesses are ready to invest in scalable live.

We’re at that rare moment where engineering requirements and business needs converge. Enter SVTA Edge with the mission focus to meet that convergence.

The key for the SVTA Edge Working Group to succeed, where others have failed, is to stay focused on what the industry needs, what is cost practical to deploy, and to avoid the temptation to solve for every interesting (forgive the pun) edge case.

That covers the why now, the why SVTA, and the what’s different this time. So let’s take a look at some of the practical steps forward that this new SVTA working group will take.

Edge Caching

Perhaps the lowest hanging fruit, the edge is a perfect place to put caches. Streaming engineers have long known the edge is a smart place to put caches, but the economics of VOD didn’t always support it. That’s because putting caches at the edge is all about lowering latency. But it’s completely acceptable to have higher latencies for VOD delivery. So from a cost perspective, it doesn’t make sense to throw more hardware to support VOD.

The focus on live, though, changes the cost equation. Increasing latency to save a little money isn’t a good trade off. It’s not as simple, however, as just adding fixed infrastructure at the edge, because the economic efficiencies of concentrated resources at the core are lost if the strategy is simply to buy, deploy, and operate duplicates of everything across the broad surface of the edge. What’s needed to make the cost equation balance, while taping the benefits of the edge, is elastic resources at the edge, which can spin up when needed and be released later. Live event caching is a great use case for edge. Live event caching doesn’t require advance pre-positioning of content, doesn’t require large amounts of storage, but does need a broad footprint of caches at the edge which can deliver low-latency streams to large numbers of viewers. When demand warrants, new containers can be spun up quickly, increasing the caching footprint to meet viewer needs, and when done those virtualized resources can be spun down and freed for other uses.

But like any containerized resource, caches are not a one-size-fits-all solution. Different kinds of video experiences often warrant different caching approaches. And when virtualizing caches, the specific use case to which the cache applies must influence the configuration of the container and the cache. Here are a few examples of different video experiences that might benefit from a specific caching strategy:

  • Massive scale, Low-Latency Live Sports Events
  • Large scale, LL-Live Sports Events
  • Medium-Small scale LL-Live Sports Events
  • DLC releases for Gaming and Devices
  • Massive non-LL Live Linear such as Major News Events
  • Regular non-LL Live Linear such as News
  • Caches with rigorous Embargo Protections for High-Value Content

Initial Project: Hosting a Specialized Edge Cache for Large Scale, Low-Latency Sports

Of course, the edge can mean a lot of things to a lot of people. But the group isn’t trying to boil the ocean. They want to make sure that they approach best practices, guidelines, or other technical documents through a very practical lens. Right now, the streaming industry is focused on live sports as one of the leading use cases driving subscriptions and viewership. But there is not yet an agreed upon set of principles or approach to the configuration and hosting of edge caches to support this use case.

In this initial project, the group will focus on defining an approach for three components of a hosted edge cache for this use case:

  • Container platform—the group will specify configuration settings for the platform, such as Docker or Kubernetes, in which the containers will be deployed
  • Hosting container—the group will look at best practices and recommendations for the container itself in which the cache is deployed.
  • Cache code—this is the caching software itself, and the group will recommend configuration, tuning, and optimization of cache settings to reflect the specific needs of the use case.

As part of this project, the group hopes to develop a reference specification for a basic hosted cache. Of course, there are several caches already used in production delivery architectures and the SVTA is not proposing to start from scratch and build a new cache. Rather, the group will look at how common caching solutions, like Varnish and NGINX, can be slimmed down and optimized for container deployment while also leveraging work the SVTA has already done.

Building On a Strong Foundation

Whether you are a proponent of Open Caching or not (we hope you are), the SVTA has done a lot of building a strong foundation for caching functionality. The group hopes to use a lot of these foundational components, like logging and capacity advertisement, in the caching reference software that they develop. Keep in mind that the group is not developing software. This initial project is a reference specification and, as such, can easily incorporate other SVTA specifications as part of the recommended approach.

Seeing the Forest for the Trees

Just developing a reference specification for the cache itself isn’t enough. The Edge Working Group recognizes that edge resources are part of a much larger picture within any network infrastructure. As such, the group is also looking at how, for each use case, other elements of edge can be addressed such as:

  • Provisioning & Control
  • Media Aware Entities
  • Scaling Responsiveness
  • Network Environment & Integration

What’s Next?

While the group will continue to tackle the different use cases for caching, there are lots of other areas that can be addressed. These might include:

  • AI/ML—how can these technologies be used at the edge specifically for streaming media?
  • Advertising—are there specific ways to virtualize ad tech stack components for deployment on the edge and would that virtualization need to take into account specific use cases?
  • Personalization—how can personalization technologies be virtualized and deployed within containers at the edge to improve the responsiveness of personalized video experiences?

Getting Involved

It’s hard not to see the benefit of this group and how it could impact the future of streaming video infrastructure deployment and management. With many high-tech operators, like streaming platforms, embracing DevOps, containers and virtualization are becoming a necessary part of the video tech stack. But the one-size-fits-all approach is just not the optimal way to use containerized tech. Misconfigurations, both at the container and orchestration level, can result in not only poor viewing experiences but even cost overruns.

Of course, the work is not going to do itself. It’s important that the group has a wide range of participating companies, from network operators to container technologies, to streaming operators, to even hardware vendors. Together, they will be able to give the industry a blueprint for employing containers and virtualized components in a manner most suited to specific use cases.

The Edge Working Group meets every other Thursday from 9AM PST / 12PM EST / 5PM GMT. The group also meets at the SVTA quarterly member meetings and will host their first, in-person meeting at the Q1 2025 Member Meeting in Tucson, Arizona.

If you are already an SVTA member, just head on over to ARO to join the mailing list. or jump in the Edge Slack channel. If you aren’t a member and this sounds like a group you should be involved in, please visit our Join the SVTA webpage and check out the membership level best suited to your company.

Glenn Deen
Comcast Distinguished Engineer at 
Jason Thibeault
CEO at  | Website

Jason is the CEO of the Streaming Video Technology Alliance, the international technical association for streaming video which brings companies from across the streaming ecosystem together to collaborate on technical solutions to delivering high-quality video at scale. In this role, he runs day-to-day operations, finances, member recruitment, strategy, and evangelizes the organization at events around the world. He is also the co-founder of a big data startup, datazoom.io. Jason is a contributing editor at Streaming Media Magazine and has written several books.

Have A Question ABout Membership?

Schedule A Meeting

Send An Email

Don’t want to schedule a face-to-face meeting just now? No problem. Simply send your membership question to info@streamingvideoalliance.org or fill out the form below and someone will get back to you as soon as possible.

"*" indicates required fields

Name*
Email*
This field is hidden when viewing the form