Blog post

The State of SDN (September 2017)

By Andrew Lerner | August 29, 2017 | 0 Comments

Here we are in the Fall of 2017 and Software-defined networking (SDN) remains a widely used and misused term that means everything and nothing at the same time. In fact, most of the networking products today marketed as SDN do not quality as SDN (note: this is based on our definition which aligns with the ONF). There was (and still is) rampant “SDN-washing” as many vendors called anything new in networking SDN (side note:  today, this is kind of like how any on-prem data center is now called a private cloud, but I digress…). SDN is an architectural model for networking, not a reference design based on a well-defined set of protocols or standards, and there are many ways to implement it. Thus, SDN is not a product, or a market.

In reality, the original architectural approach behind SDN has achieved limited acceptance despite high-interest (we’ve received 1,700+ SDN-related inquiries since 2016). However we estimate, there are less than 2,000 deployed networks that meet the architectural requirements for SDN. Not surprisingly, it is pretty close to the trough of disillusionment in our 2017 Enterprise networking hype cycle. There are several reasons for this including:

  • the conservative nature of networking buyers
  • concerns of stability and scalability of SDN
  • technological immaturity of the standards
  • lack of immediate business drivers
  • difficulty in putting a value on agility and innovation to prove ROI
  • market power of incumbent vendors that had a vested interest in protecting their market position.

That said, SDN had an extremely positive impact on the networking industry because it changed the conversation in networking. SDN shifted the perception of value from hardware to software. Further, it showed the value of open, northbound controller APIs in enabling a rich ecosystem of partners. Without SDN, disaggregation of the switch hardware and NOS and the emergence SD-WAN would have been far less likely to occur. At this point, in the battle of SDN versus legacy networking, it’s now time to declare victory and move on, not because SDN won or lost, but because it drove needed changes to the industry.

This (and more) is covered in our latest research on the State of SDN

Summary: SDN promised a new era in networking, but left many buyers confused about its value and how to implement. To deliver on SDN’s promise, I&O leaders responsible for networking must refocus away from SDN architecture, and onto achieving agility, software capabilities and hardware disaggregation.

Regards, Andrew

PS – Happy 50th Anniversary to my parents, who will never read this, but it feels wrong to publish anything on 08/29/2017 without noting it!

Leave a Comment