Feature Story

Fog Computing: Innovation Beneath the Cloud

As the Internet of Everything heats up, Cisco engineers propose a new networking, compute, and storage paradigm that extends to the edge of the network.

Early on November 3, 2007, heavy fog caused a traffic pile-up that eventually included 108 passenger vehicles and 18 big-rig trucks on a busy highway in California's San Joaquin Valley.

Fog may provide life-sustaining moisture for California's ancient sequoia trees, but it has been the bane of transportation ever since the first stage coaches began clattering up and down the Golden State in the mid-1800s.

Some 150 years later, a new kind of fog emanating out of Silicon Valley could make traffic pile-ups worldwide a thing of the past.

Fog nodes, with compute, storage, and networking capabilities, extend from the very edge of the network to the Cloud. Fog nodes are not homogeneous. Usually small and ruggedized at the very edge, they become more powerful at the higher layers.

When the truck ahead of you brakes unexpectedly, you want your car's sensors to get that information from the other vehicle instantly. And some data – such as available parking spaces at the mall – might have no lasting value and never need to go to the cloud at all.

"Fog is an expansion of the Cloud paradigm," says Technical Leader Rodolfo Milito. one of Cisco's thought leaders in fog computing, "It's similar to cloud but closer to the ground. Fog computing architecture extends the cloud out into the real world, the physical world of things." Fog complements the Cloud, addressing emerging IoT applications that are geo-distributed, require very low latency, or fast mobility support. Fog computing would support sensors (which typically measure, detect, and collect data) and actuators – which are devices that can perform a physical action such as closing a valve, moving the arms of a robot, or exercising the brakes in a car.

Fog could also take a burden off the network. As 50 billion objects become connected worldwide by 2020, it will not make sense to handle everything in the cloud. Distributed apps and edge-computing devices need distributed resources. Fog brings computation to the data. Low-power devices, close to the edge of the network, can deliver real-time response.

Smart grid is a good example. "You cannot drive the smart grid from the cloud. You need to be able to react at the periphery, quickly," Milito says.

Fog could also handle critical data in places where access to the cloud is difficult, slow, or expensive. On a deep-sea oil rig, the fog device would monitor vital sensors in real time but transmit only select or aggregated data over the satellite connection.

Fog may seem intuitively obvious to some but it's extremely disruptive to traditional networking. Distinguished Engineer Don Banks is working with the fog team to develop the architecture for fog computing, and he says that introducing a new networking paradigm is the biggest challenge.

"Fog computing devices are multi-use," Banks says. "They provide computing, storage, and networking in one device. And they're multi-tenant, which means that one device could support several different organizations and companies. That requires a highly virtualized and secure system architecture."

"The Internet of Everything is changing how we interact with the real world," Milito says. "Things that were totally disconnected from the Internet before, such as cars, are now merging onto it. But as we go from one billion endpoints to one trillion endpoints worldwide, that creates not only a real scalability problem but the challenge of dealing with complex clusters of endpoints – what we call 'rich systems' – rather than dealing with individual endpoints. Fog's hardware infrastructure and software platform helps solve that."