CDN history has been a roller coaster for the industry. From single CDNs to multi-CDNs to using just a few, dealing with more than one delivery provider is complicated, requiring bespoke configurations and switching approaches. Can capacity reservations simplify the multi-CDN complexity?
Delivering content efficiently is about more than having enough bandwidth. Geographic proximity to users, accurate capacity planning, and visibility into delivery infrastructure all play critical roles in ensuring a high-quality experience. As content providers face growing demand and increasingly distributed audiences, new standards and marketplace models are emerging to help them identify, reserve, and access the right capacity in the right locations at the right time.
Before we dig into capacity reservations (and how geography adds another dimension of complexity), let’s talk about why caches located closer to the viewer are so critical to content delivery. When a user makes a content request, that request is handled by a number of different parties including the viewer’s ISP, and other network operators, on the way to retrieving the content and returning it. The time it takes is called the Round Trip Time, or RTT. As you can probably surmise, the longer that RTT is, the worse user experience. But if that content is located more closely, say in the edge of the ISP network or at the edge of a network peered with the ISP (meaning both networks are physically located in the same data center and, possibly, even cross-connected), the RTT time is significantly reduced. This, in turn, helps promote a better user experience.
While having edge caches is preferred, that is not always an option. When you consider Tier 2 and Tier 3 ISPs, for example, who may service several hundred thousand people down to a few thousand, they often don’t have the expertise or capital to host their own caches nor are they located in proximity to an Internet Exchange or other peering location. That means their users’ RTT is usually longer. And when you consider just the United States, the number of users located outside of Tier 1 ISPs numbers around 50 million. But the impact of that longer RTT is more than just unhappy users — it can also impact the content owner. When a rural user has a poor experience with a streaming platform, for instance, they may terminate their subscription even though it’s not the streaming platform itself that is the problem but the poor Quality of Experience (QoE) resulting from longer RTT and lower bitrates.
Content providers need to reserve capacity when they have big events. In streaming video, this might be a big live event. In gaming, this might be a highly anticipated release or update. So when these big events are coming up, the content provider often works closely with their CDN partners to ensure that the network has the available capacity in the regions where they need it. This usually involves a targeted amount, say X Gbps, and a burst amount just in case the traffic estimates are below what actually happens. If capacity is determined by the aggregate availability of the entire caching pool (it’s not just the size of the pipe, but the amount of throughput that each NIC on each cache can handle as well as the link throughput available from each peering partner in the exchange) but there aren’t edge caches available deep within the targeted geographies, then the available capacity, although true, isn’t accurate in terms of providing the highest level of QoE. Sure, a given CDN might “cover” that geographic area, but the servers are a hundred miles away. This can often be the case with Tier 3 ISPs where their upstream capacity provider, a Tier 2 or Tier 1 ISP, is located a considerable distance away…and where the other delivery networks, like CDNs, are peered.
So what happens when a content provider is trying to deliver the best possible experience into a very specific geography and their delivery partners don’t have physical infrastructure there? You guessed it — bad user experience, poor QoE, and potential business impact. This is why capacity reservations add such a layer of complexity to content delivery. Content providers have to rely on the word of the capacity network that their footprint is capable of delivering a great experience in those areas.
The Streaming Video Technology Alliance (SVTA) has created a family of specifications around what’s called Open Caching. This is a technology framework of APIs which allows the sharing of information between CDNs and the content providers that use them. One of those APIs, Capacity Insights, allows a delivery network to “broadcast” its available capacity. When this data is coupled with zipcode distribution of infrastructure (caches), the content provider can actually get a realistic picture of the actual available capacity at any given moment in time. This kind of programmatic insight can help content providers not only plan more accurately for big events, but do so in near real-time. So rather than coordinating with a CDN weeks or months in advance, the content provider can simply ping the Capacity Insights API of the ASNs they want to work with to get a true understanding of capacity in any given area served by that delivery network.
While accessing near real-time capacity is a great way to reduce the complexity of delivering content, it may also reveal something — gaps in geographic coverage. As was pointed out already, not all CDNs are created equally in every geographic region. Some CDNs might be in the Internet Exchanges (IXs), peering directly with major ISPs. Others may have partnerships with specific data centers, giving them a “super PoP” capable of handling a lot of requests from surrounding areas. But there are also geographic regions don’t have any content delivery physical infrastructure (caches) and must rely on the networks between users, ISPs, and CDNs to ensure a great user experience with requested content. But there is more capacity that may be available. Satellite ISPs, mobile operators, and even broadcast locations which are ATSC 3.0 capable, might have capacity that can be used by a content provider to deliver to those hard-to-reach geographies. Unfortunately, those alternative capacity providers might not be available and, even if they were, it would require more bi-lateral business agreements, complicating content delivery even further.
Blockcast, in conjunction with the SVTA and Vecima Networks, has created a not-for-profit industry foundation called the Open Capacity Marketplace (OCM) to address many of the complexities of content delivery including reservations and access to non-traditional capacity. Based on the SVTA’s Open Caching specifications as well as the IETF’s CDNi work, this marketplace is a technology platform that connects content providers with capacity suppliers in a permissionless environment — requests for capacity (based on type and geography) are submitted and replied to in a seamless fashion. What’s more, the bi-lateral business relationship is encapsulated in a Web3 Smart Contract, delivered bits are stored on chain (for immutability purposes), and settlement can happen automatically by comparing the delivery data with terms in the Smart Contract (the SLA). In addition, Blockcast can deploy agents (as part of their end-user focused Web3 CDN) that can help verify that capacity independently.
By leveraging the OC and CDNi specifications within the OCM, content providers can get accurate capacity reservations (based on what is actually being communicated by the delivery network in specific regions), reduce the friction of managing a lot of bi-lateral business relationships with a lot of delivery networks, and even get access to that non-traditional capacity.
Are you a content provider or a capacity supplier? Blockcast and Vecima will be showcasing a demo of the OCM at the Vecima booth (Hall 1, stand B15) on Saturday at 4:30pm. Drinks and light refreshments will be available. After the demo, attendees will have a chance to talk with representatives from Blockcast, Vecima, and the SVTA about the OCM.
If you are interested in attending, simply visit the IBC 2025 Demo section on the Open Capacity Marketplace website and fill out the form to indicate your interest in attending.
Edge caches store content closer to end users, reducing the round-trip time (RTT) required to retrieve content. Lower RTT typically results in faster load times, reduced buffering, higher video quality, and a better overall user experience. By placing content near viewers, content providers can improve performance without requiring users to traverse long network paths.
Geography plays a major role in content delivery because not all internet service providers (ISPs) have nearby CDN infrastructure. Many Tier 2 and Tier 3 ISPs are located far from major internet exchanges and CDN points of presence (PoPs), resulting in longer RTTs and potentially lower quality of experience (QoE). For content providers, this can lead to customer dissatisfaction and even subscriber churn despite the content service itself functioning properly.
Capacity reservations allow content providers to secure sufficient network resources for anticipated spikes in demand, such as major live sporting events, software releases, or game updates. By reserving bandwidth and delivery capacity in advance, content providers can help ensure their audiences receive a consistent experience even during periods of unusually high traffic.
The SVTA Open Caching framework includes a Capacity Insights API that allows delivery networks to communicate available capacity programmatically. When combined with information about cache locations and infrastructure distribution, content providers can gain near real-time visibility into actual available capacity within specific regions, helping them make more informed delivery decisions.
The Open Capacity Marketplace (OCM) is a platform created by Blockcast, the SVTA, and Vecima Networks that connects content providers with traditional and non-traditional capacity suppliers. Using Open Caching and CDNi standards, the OCM enables automated capacity requests, geographic-specific reservations, smart contract-based agreements, automated settlement, and access to alternative delivery resources such as mobile networks, satellite operators, and ATSC 3.0 broadcasters. This helps reduce operational complexity while improving access to delivery capacity.
.png)