Converging in the Fog
By Mychal McCabe
A reasonable person could be forgiven for thinking that the distinction between centralized and distributed computing models is beginning to come to an end, with Fog architectures bringing computation, communication, control and storage tantalizingly close to edge devices.
Hype around a handful of technologies is fueling the fire, with each viewed as a phenomena in a distinct market or market segment rather than the set of innovations that will accelerate convergence of the network, data, and control into a pervasive Internet of Things.
IoT is the most visible thanks to popular consumer devices, broad applicability of the concepts across market segments, and the sheer opportunity for everyone along the value chain. SDE or Software Defined Everything is a catchall for the variety of technology domains where software is being decoupled from specialized hardware: Software Define Networking (SDN), Software Defined Data Center (SDDC), and Software Defined Storage (SDS). Finally, Network Functions Virtualization (NFV) virtualizes network services that are carried out today by proprietary, dedicated hardware.
From an IT point of view, Software Defined Everything addresses everything so long as it happens to be in a datacenter or the enterprise. From an OT point of view, there are specific applications where the business benefits of software defined architectures are compelling enough to prompt something that in the early days of our enthusiasm for the Internet of Things seemed unlikely: a massive rip and replace of brownfield infrastructure resulting in a connected and more intelligent edge.
The big cloud players see this increase in intelligence and connectivity as an opportunity for their analytics capabilities to migrate further toward the edge (simple analytics on a single device, more complex multi-device analytics, and machine learning) while at the same time opening up the possibility of specific workloads and some classes of control logic migrating from the edge to the cloud. Not surprisingly incumbents in the OT domain are leveraging their expertise and footprint at the edge to move toward the cloud.
Fittingly the current proving ground for application of these technologies, which relate to one another in ways that are only beginning to become clear, is what our customers and our partners are increasingly referring to as The Fog.
The fixed function nature of embedded systems –software “cut to fit” for specialized hardware, built from the ground up to do one thing– results in exceptional performance and reliability that cannot be easily replicated when migrating control logic to the cloud. The trade-off comes in the form of added complexity, as addressed in the OpenFog Consortium FAQ:
“While adding fog computing to an IoT network would seem to add complexity, that complexity is sometimes necessary. In certain use cases, fog computing solves the inadequacies of cloud-only models, which have serious latency, network bandwidth, geographic focus, plus reliability and security challenges.”
This is what the early days of the Internet of Things looks like: Aggressive application of proven, disruptive IT technologies and lifecycle approaches –virtualization, cloud, analytics, DevOps, containers, etc.– applied to OT: edge computing applications that have historically been the domain of embedded systems.