An edge computing architecture upends the data center

BACKGROUND IMAGE: carloscastilla/Fotolia

Through all the fog, edge computing emerges

It's unsurprising that there's so little agreement about how best to deploy data center technologies in a widely distributed way. Indeed, there isn't much consensus even about what constitutes an edge data center.

What is clear is the inclination to move ahead in pursuit of what are described as fog or edge computing opportunities. The area at the network's edge is simply too inviting to not proceed. The edge is where the internet of things (IoT), mobile computing and other time-sensitive applications will either thrive or languish.

An edge data center is all about putting computing and storage infrastructure out there in the world, close to users, where it can be productive in something like real time. This decentralized approach is what is meant by edge computing architecture. Local placement of resources becomes imperative when the conversation turns to IoT. For serious IoT implementations, analytics capacity needs to be nearby. Processing data over a vast distance just isn't going to get it done when near-instantaneous reactions are expected.

As contributor Alan Earls reports in this handbook's first article, those working on the edge will face some serious challenges while realizing powerful advantages. Latency, for one thing, should improve considerably. The right technologies placed opportunistically will speed the exchange of data, thereby enabling location-based services that would otherwise be problematic, if not impossible. Retailers, in particular, covet the power to serve up deals directly to customers who are in or near one of their stores. That real-time engagement with shoppers is dependent on effective edge deployments.

Potentially profitable fog or edge computing uses will require investment in products, of course, as well as management skill. IT decision-makers need to figure out how to best run applications at the edge and how to handle data retention and management. The biggest concerns with an edge computing architecture are -- and probably always will be -- about the security of data and devices that are so widely distributed.