Lately the concept of “fog computing” has been showing up more and more often in the press and blogosphere. Introduced by Cisco, fog computing is a distributed architecture in which computing is moved to the edge of the network. Due to the massive growth of data coming from a huge number of sensors and devices in the Internet of Things (IoT), it’s becoming incredibly challenging to move all this data to the cloud – which requires moving it over networks to a remote cloud data center. Since the goal of gathering all this IoT data is to be able to analyze it in near or real-time and to look for trends and anomalies, a more distributed architecture is needed. Enter the fog.
Ignoring the unfortunate name choice, fog computing definitely makes sense. Given the surge of data at the network edge, there is a need for intelligent processing of this data before some or all of it makes its way to a cloud “hub” for deeper analysis. In practice what this means, and what Cisco is promoting, is to put smart devices at the network edge so that some processing can be done locally and decisions can be made on which data needs to be moved. The result is much less movement of data overall, and much less back and forth to get the results in a timely way.
But fog computing is really pinpointing one aspect of a much broader issue. IoT data, while growing exponentially, is still an emerging market built around a limited set of applications and use cases. In the meantime, what about all the existing enterprise applications? As hybrid cloud models become widespread, the movement of data between enterprises and the cloud is steadily increasing, straining networks and the ability of IT teams to determine what should stay local vs what should move to the cloud.
What’s needed today, and more widely, is an approach that encompasses the concepts of fog computing but applies them to the mainstream business challenges that affect all enterprises. As IT transitions to becoming a service provider for internal business units and users, “IT as a Service” will include traditional infrastructure models in addition to external SaaS, cloud and managed services. So how is all this data going to be managed? And how will IT figure out which data should stay at the edge? The balance between enterprise data centers and regional cloud “hubs” is a challenge IT teams are just beginning to face, but it’s at the heart of the success of hybrid models and cloud adoption overall.