Cloud computing, data mobility and the Internet of Things (IoT) are pushing more of the enterprise workload onto the Internet, which naturally leads to increased latency, service interruptions and general frustration over the state of modern digital communications.
But while telcos and other carriers have done a good job at maintaining available bandwidth, satisfying the growing complexity of data traffic is another matter. Ideally, what is needed is a way to get data storage and processing closer to users so they can get what they want without having to traverse miles of wired and wireless infrastructure.
This is why edge computing has taken on greater urgency of late. In one move, the enterprise can improve service to users, lower its wide area networking costs and perhaps create new revenue-generating services to boot.
The trick, says Amy Newman for eWeek, is to maximize the value of edge computing without adding unnecessary complexity to distributed data environments. At the moment, edge systems serve three primary functions: a local storage center for data-heavy content like hi-def video; an aggregation and control point for connected devices; and an on-prem application and processing tool for replicated cloud services. Each function will require different configuration options, of course, ranging from small appliances to larger systems that begin to resemble regional data centers. But in all cases, the intent is the same: to keep as much data as possible from jamming wide area network infrastructure and centralized data resources.
For organizations looking to implement IoT solutions (and that should be just about everyone by now), it is difficult to see how this can be done effectively without edge computing, says Patrick McGarry, vice president of engineering at Big Data solutions provider Ryft. Much of the IoT data load will be ephemeral in nature – that is, it will only be of real value for a very short time. So that means it has to be captured, analyzed and acted upon very quickly. Think of a retail environment that can analyze customer traffic, buying history and multiple other data sets to target sales opportunities at in-store shoppers, or an oil-drilling operation that requires real-time results to avoid potential disasters. Neither of these scenarios would be possible if data had to traverse the Internet to a central processing facility and then back to the field where the decision makers are.
Naturally, solution providers are eager to get in on what will likely be substantial growth market going forward. Dell recently added a company called flowthings.io to its IoT Solutions Partner Program, giving it access to the Edge Execution Engine that combines edge gateway technology with real-time data orchestration to support IoT workloads. The platform functions on multiple edge devices, such as the Dell Edge Gateway 5000, to handle processing, filtering and routing of edge data, while a self-contained, full executable edge extension agent for Linux and Darwin UNIX platforms coordinates data flows between the cloud and the edge. As well, it provides a range of drag-and-drop development tools and APIs to execute instructions at the edge.
Since it is quite likely that edge processing will proliferate across enterprise infrastructure, the devices themselves will have to be compact, fairly intelligent, and capable of remote service and upgrades. The groundwork for this is already underway with new compact board- and device-level solutions from companies like Supermicro. The X10SDV motherboard, for instance, now supports 16-core Xeon D-1500 processors to support scale-out hot and cold storage on converged infrastructure. The company has also released 1U server, tower and PC devices with low-power mesh networking capabilities for edge-to-cloud applications.
While it may seem for a while that edge computing is the wave of the future in IT, it’s important to remember that it is not intended to replace centralized infrastructure but to share the burden of increasingly heavy data loads. In fact, a good chunk of IoT data is only valuable on the edge, while the rest may only be useful for more strategic applications in the cloud.
The most forward-leaning edge solutions, then, will be the ones smart enough to know what to keep and what to pass on.
Arthur Cole writes about infrastructure for IT Business Edge. Cole has been covering the high-tech media and computing industries for more than 20 years, having served as editor of TV Technology, Video Technology News, Internet News and Multimedia Weekly. His contributions have appeared in Communications Today and Enterprise Networking Planet and as web content for numerous high-tech clients like TwinStrata and Carpathia. Follow Art on Twitter @acole602.