• Innovation doesn’t mean the technology has to be net new. It just has to significantly change a direction.
• SDN has had and will continue to have a significant impact on enterprises and vendors for years to come.
SDN is many things to many people and the answer to the question “What is SDN?” will vary depending on who you ask. In fact, I’d say that SDN has become so overloaded that it has lost its meaning. But unlike my friend Greg Ferro, who thinks SDN is not an innovation, it’s iteration, I think SDN is innovative—or at least parts are—but more importantly, SDN unlocks innovation.
Sure, we can look at technical components of SDN like automation and overlays and compare them to similar technologies from the past and conclude these new things are not innovative. By that hindsight, nothing in the last 50 years is innovative. SDN is more than just automation and overlay. To pull an old saw, the sum of SDN is greater than the parts. What I think is innovative about SDN is the new approaches that it is bringing to the table – enabling more intelligent networking between nodes in the network, among applications, and between locations. Approaches like:
• Microsegmentation or the ability to limit access between nodes regardless of the network topology on an extremely granular level with operational ease.
• Performance based forwarding in LAN hardware by evaluating various counters in network hardware and responding by selecting the best path.
• SD-WAN brings together heretofore disparate products and technologies into a cohesive whole, greatly simplifying remote networking with path selection, cloud service insertion, policy enforcement, and management by preference.
• User driven networking—still rather nascent but certainly possible in the LAN and definitely coming in the WAN—where the local or wide area network is provisioned and configured with little to no user involvement.
(I can do this all day.)
It’s not that these capabilities didn’t exist prior to SDN. It’s that these capabilities weren’t readily available prior to SDN and that’s the innovative part—the making of these capabilities and many more readily available through packaging, integration, and automation. For example, with SD-WAN, enterprises can replace five technologies implemented on separate hardware in the branch and still make it to the pub for beer o’clock. Could enterprise IT have supported all the functions of SD-WAN products prior to this? Sure they could and they did. But SD-WAN isn’t merely iterative because it consolidated functions into a single appliance. SD-WAN is innovative in how the technologies are integrated unlocking new capabilities.
I think as the use cases for SDN in the enterprise develops and vendors and open source projects rush to address those use cases, we’ll see more net new capabilities that will be innovative, whether or not they bear some passing resemblance to the past.