Two presentation at Gartner’s recent IT Operations & Strategy Summit in Orlando provided some interesting perspective on this deceptively complex question: what is the cloud edge? One was by Oracle Dyn GM & VP of Development Scott Hilton, who presented on IaaS and the edge, the other was by Gartner VP and Distinguished Analyst Thomas Bittman in his keynote titled “The Journey to the Mix: Private Cloud, Public Cloud and Edge.” While these presentations came from unique perspectives, they both point to the need for organizations to pay close attention to the edge in the age of cloud.
As cloud computing has gained prominence, we’ve seen plenty of analogies comparing cloud to mainframe computing. The notion of large, centralized compute (and relatively dumb endpoints) made this analogy apt, but fall short in reality. Bittman makes this point by pointing to the requirements driving the need for compute at the edge: applications like autonomous vehicles and augmented reality. As a cloud DNS infrastructure and internet performance management company, we’ve long been obsessed with the experience users have connecting from the edge. We know from years of monitoring the internet at scale the numerous things that go bump in the night that contribute to the volatility of the internet. It’s from this perspective that we introduced the concept of internet performance management and DNS with “intelligent response.”
Both visions for the edge (Dyn’s and Gartner’s) share a notion of an “intelligent edge.” Bittman spoke to the edge requirements for local intelligence, data, interaction, combined with low latency and real-time processing. Image the use case of autonomous vehicles, equipment or industrial robots interacting with their environment based on ever-changing conditions (see diagram below).
The Dyn notion of intelligent response is a version of an intelligent edge at the infrastructure / network; an edge (i.e. DNS network) that’s infrastructure-aware and able to intelligently route traffic to the optimal end-point based on policy and real-time performance data.
Bittman’s focus on compute-intensive edge applications, particularly IOT and consumer augmented reality, points to a future architecture that’s much more complex and much more hybrid. Some compute will happen locally (in the car), while some will occur in the cloud core (big data analysis, telemetry, etc.). Organizations need an edge strategy that can accommodate for this burgeoning heterogeneity.
As for defining the edge, Gartner’s Tom Bittman provided a clear and straightforward definition as the point where people and devices/things connect to the network. Because of our focus on intelligent response, we’re positing a slightly more nuanced definition of the edge: The enterprise edge is where the enterprise has control over user access or services. This definition suggests that your edge could, should and increasingly must do more to deliver on both the promise of the cloud core and its increasingly complex edge.
Interested in learning more? Scott Hilton recently reprised his Gartner presentation as a webinar, available on-demand on our hub.