As I read articles and blogs over the past few weeks, I was astounded to see that, all of a sudden, a number of authors were claiming that SDN is dead. They cited the fact that vendors throughout the industry all have applied an SDN label to their varying offerings and consequently claimed that SDN has a become a meaningless acronym! Interestingly enough, these same authors then went on to call for SDN 2.0…as if history would not repeat itself and as if SDN 2.0 would be precisely defined, compared to SDN.
The fact is that SDN holds real promise and the industry is starting to deliver on it, albeit in a staged manner and in different ways. This is to be expected and hoped for at this stage of the evolution and, if nothing else, shows the power of SDN. It is also good to remind ourselves that SDN is a concept, vendor product and implementations will by definition differ depending on environment, business challenge, and implementation. And yes, SDN is hyped, just as many other technologies that have preceded it. However, customers have always managed to separate the wheat from the chaff, and as we move to real-world implementations, we are starting to see solutions delivering on the SDN promise and some concepts prevailing over others.
As we move toward broader adoption, there are two major questions that customers across different market segments are asking consistently and vendors need to make sure they are ready to engage with and respond to.
1. How can we identify business problems SDN can realistically address quickly and can SDN / SDx help address the issue in a phased manner?
This is a loaded but very valid question with a number of implied statements, specifically that customers at large can’t afford “big bang” infrastructure changes. Vendors need to start to engage with customers to really understand their business issues and take the time to build a proposition to address them with the underlying technology and automation – a very different and far more engaged approach than what technology vendors are used to.
2. How can SDN be applied to help reduce cost?
Again, a very valid and pressing question in many environments where IT departments are asked to do more with less. The SDN promise in its purest form with whiteboxes and a singe simple controller offers just that, right? Wrong! Reality is that, with very few exceptions, most networking vendors’ equipment has become significantly cheaper over the past years, as they are all buying from the same original equipment manufacturers (ODMs). Unless you are in the hyperscale data center business, this doesn’t differ that much from a Dell approach anymore. And secondly, it is widely known and has been proven by numerous pieces of research that the initial capital outlay represents merely a fraction of the total lifecycle spend associated, opex being the largest portion.
So how does this answer the question, you may ask? Well, as stated at the outset and without drilling down into the technological intricacies, the fundamental promise of SDN is to provide cost efficiency, provide predictability, reduce time to service and improve the rate of innovation by applying automation, virtualization, programmability, simplicity and open standards as well as a surrounding ecosystem. The logical consequence of this promise, combined with the innovative nature driven by the community and the framework as well as the different points of departure, is that no single SDN implementation will look like the other.
SDN, if done right, provides an abstraction layer that can be programed in an agile fashion without having to care too much about the underlying infrastructure and, at the same time, removes some of the legacy complexity. It provides the simplicity we need to allow innovation to flow effectively and without limitations to result in a multitude of different SDN-based, business-relevant solutions.
I would therefore argue that SDN 2.0 is a meaningless term. What we need is a realization and an acknowledgement that SDN is not necessarily equal to SDN and we need real and relevant SDN deployments in business critical environments, focused on delivering real business value and not yet another acronym. Customers are asking the right questions and should push their vendors for answers. So let’s stop arguing semantics and focus on what really matters: delivering innovative SDN implementations that solve real problems today.