In the new world of always-on technological demands, high-speed applications and highly mobile data requirements, it’s hard to escape the term ‘edge.’ By now, most know the edge as the very fringes of the network - the areas closest to the user or to the data’s point of origin. We know edge computing’s advantages, most notable of which are lowered latency and heightened network performance, and which applications are driving us toward that edge (think 5G, artificial intelligence, cloud gaming, IoT and others). But although the benefits of the edge have been made plain, the infrastructure and processes that make the edge tick can still be shrouded in mystery.
To leverage edge computing to its fullest extent, understanding the foundations of the network’s edge is key. In particular, understanding the role of peering in the edge’s evolution (and the applications that flourish there) is becoming crucial for successful digital transformations. As data must become available at more endpoints, reach more platforms and move through the network more quickly, it’s clear that even leveraging a single edge data center may no longer cut it. So, it’s time we dive in and explore how peering is the heart of the edge.
Old world interconnection - and why it fails modern use cases
Peering, despite its increasingly pivotal role in contemporary networking, is founded on a simple idea: connecting one network to another. Connecting networks serves to expand the reach of both, allowing data to traverse the entirety of the combined network, and enabling participants to get data to a growing footprint and a better range of destinations. However, its advantages lie in one key difference from typical transit connections (an old standby in the IT world).
Yesterday’s methods of network interconnection involve roping in a third party transit provider - a paid upstream network participant that will carry the traffic to the wider internet on behalf of the customer. Aside from the additional costs this method incurs, there can be an array of potentially damaging factors at hand. In this case, once the traffic is handed off by the enterprise to the transit provider, they relinquish control over how that data moves. In extreme cases, this could mean the data is going to the opposite side of the country before reaching its final desired destination, adding latency and impacting network performance.
In an edge-driven era, latency and speed are where it counts. So, risking lackluster performance by relinquishing control over data flows is no longer an option for businesses looking to remain competitive. This is where peering works its connectivity magic. With peering, networks connect to each other directly, avoiding middlemen and additional costs while taking back control over their data. With peering, there are no additional charges beyond the costs associated with powering the necessary routers and switches, and the exchange becomes a symbiosis for all involved parties.
How the peering edge has taken shape
When latency is reduced and speed is preserved with peering across a growing number of internet exchanges (the locales where direct interconnections take place), we see a foundational element of edge computing take shape. In many ways, the network edge and the network of peering locations and networks have become the same, exhibiting a direct relationship in growth — peering thrives on the demand of edge and mobile applications, and as edge computing continues to grow, peering does the same. Herein lies the origin of the peering edge.
Now, direct interconnection is no longer transit’s backup plan, or just another alternative for procuring connectivity. Instead, it’s become the only option in many cases for increasing network reach, preserving performance and ensuring applications can function the way businesses need them to. So, the question for the enterprise now becomes how to incorporate internet exchanges into their edge growth strategy, because although all peering offers advantages, not all exchanges are made the same.
Cultivating the right connectivity strategy
Identifying the perfect peering partners and locations is a vital consideration in any digital transformation, especially those supporting edge-centric demands. When asking themselves how traffic flows can be optimized for individual needs, there are a few things enterprises must consider. The first is probably obvious - location. A peering exchange must reside in an area that is convenient for the business, and as peering footprints grow, additional exchanges should align with the geographical areas the business needs to connect into most.
Beyond the geographical location of an internet exchange, any peering locale’s most important asset is its ecosystem. Since peering is founded on relationships and connections, having the greatest number of connectivity opportunities in one place will serve an enterprise well. With more peers come more options for redundancy and reach, and as these ecosystems grow, they acquire a kind of gravity. The larger the ecosystem is, the more networks are likely to join it, which makes this a type of self-sustaining growth and momentum - ultimately setting peers at that location up for long-term connectivity success.
Peering policies can also be good to consider when checking out different exchanges, as they can impact the peering experience with different levels of selectivity or restriction across the ecosystem. Policy considerations such as the exact method of interconnection at each exchange, and how they compare between available locations, can be checked out through tools like PeeringDB.
At its core, peering is partnership. With the right partnerships in place, whether with other networks or a peering provider, edge initiatives can be greatly empowered, and digital growth can be set up well for the future’s demands. Certainly, as the edge continues to build its reputation as the promised land of IT innovation, applications and digital transformations will only continue to align this edge with direct interconnection.
