There seems to be a general belief that software-defined networking (SDN) will eventually redefine today’s decades-old, hardware-centric architectures. However, even as we look forward to the potential of SDN, it’s hard to accept the reality if you think about SDN’s past.
Although Gartner placed SDN at “rock bottom” on its 2015 Networking Hype Cycle and has it “firmly entrenched” in the trough of disillusionment for 2016, there’s hope in the narrative.
If you look at the overall story arc for SDN, you’ll see that what was once a lot of hype is now (slowly) becoming reality.
It should come as no surprise that Google’s internal network is one of the biggest (if not the biggest) in the world. More than that, though, Google’s private network has grown at rates faster than the internet itself at times.
Google realized in 2008 that they wouldn’t be able to sustain such massive growth with traditional means of scaling hardware-centric networks. They turned to SDN in response, implementing its first iteration in 2010.
According to Google’s tech lead for networking, Amin Vahdat, “Through centralized traffic engineering and quality of service differentiation, we’ve managed to distinguish high-value traffic from bulk traffic that’s not nearly as latency-sensitive. That has made it possible to run many of our links at near 100% utilization levels.”
Google’s SDN success and the promises of centralized network control and visibility led to the early hype surrounding software-defined networks. Unfortunately, there are only a handful of companies in the world with the resources necessary to achieve Google’s SDN reality, leaving most businesses trying to implement SDN to face significant challenges.
There was a time when processing speeds and performance were the key concerns when trying to replace hardware-centric networks with SDN. However, innovation surrounding multicore processing and commoditized hardware are pushing us closer to an SDN reality.
Despite improvements in processing performance, there are reasons why SDN still sits in Gartner’s trough of disillusionment and largescale deployments are rare. There are two persistent challenges that SDN innovators must overcome to advance the SDN narrative:
Industry expert Jennifer Rexford has said that “if SDN is going to prove successful in a much broader context, it’s going to be because there are reusable platforms available, along with the ability to build applications on top of those platforms.”
This may not be today’s reality, but networking demands are pushing innovation at a more rapid pace.
While the hype has died down due to SDN disappointment, the future of networking still lies in SDN. The market is set to grow exponentially over the next 4 years as industry leaders focus on the specific challenges still facing widespread SDN adoption.
A year or two ago, it might have been acceptable to ignore SDN as technology continually failed to realize its lofty promises. Now that expectations have been tempered, SDN can be taken more seriously.
AT&T’s Senior Executive Vice President, John Donovan, once recognized the need for revitalized IT training to prepare for a software-defined future. It’s no secret that overhauling network architecture standards that have stood for decades will be a challenge, so you have to prepare yourself now.
There are multiple roadblocks to meeting growing network demands including, complex management, limited visibility in tool silos, vendor lock-ins, and more. Adapting to the emerging SDN reality will require solutions built for tomorrow’s network speeds and capabilities and ensuring that visibility is built into the architecture.
Learn more about the benefits of SDN in 40G/100G networks, as well as the importance of architecting data centers with visibility for both the data plane and the control plane.