Open Networking Is Gaining Steam, But Challenges Persist


Open Networking User Group

Earlier this month I attended the Open Networking User Group (ONUG) event in NYC, the bi-annual opportunity to hear from customers about how networking is still not meeting their needs and what vendors need to be doing to address the situation. While the networking kingpins like Cisco Systems, Juniper Networks and Alcatel-Lucent Enterprise own the network landscape, it is clear that the latest innovations are coming from two places: smaller startups and open networking. Both of these have an opportunity to challenge the status quo; the startups in taking business because they can move faster than the incumbents, and open networking because it challenges the proprietary lock-in that has been the hallmark of the traditional networking business.

In talking to customers, I hear over and over that they are not satisfied with the current options. Networking is too costly, too complex, it holds the business back and prevents companies from taking advantage of new opportunities. When businesses are trying to move “at the speed of the cloud”, having networking that takes months to provision is simply not going to cut it in today’s rapidly changing world. Customers want more speed and less complexity. But the reason that things are slow and complex is that they have been built over years of “new” hardware that promised change but never really delivered. Now, as startups challenge the business with new innovation that can literally rewrite the book on how companies tie their resources together and communicate, delivering exactly what customers want, why aren’t these startups printing money at this point? And if these new standards are so much better than the old way of doing business, why aren’t people abandoning the old way of doing things and embracing the future?

The simple answers are money, infrastructure and risk–but these are only speed bumps, not roadblocks.

Money is a tough one because when you talk to these customers, they tell you over and over that they are spending too much on equipment today and getting too little. But yet they persist. Often, as with some of the large core switching, it comes down to amortization. While non-traditional challengers like Dell or Hewlett-Packard Enterprise would love to swoop in and pick off some marquee customers, the reality is that much of the existing equipment running networks today has not been amortized. And until it is fully depreciated, it is difficult to consider changes. There are a lot of people driving around in cars that they hate today, but until they are paid off, you won’t see them in a shiny new one. Over time though, we will see changes happen; while it may look like a slow trickle today, there is every expectation that fully amortized core equipment brings more options to customers as the vendor gravity changes.

Infrastructure is also an issue. When you have a datacenter full of company X’s equipment and you need to expand or replace something, continuing down the same path is usually the path of least resistance. Too often we hear that continuing down the same path–while not the direction that the customer wants to go–is the best solution when the business is pushing for more network services and faster deployment. It’s simply too difficult to change that engine if the plane is always in flight. The silver lining here is that as new datacenters are being built, companies are starting to second source equipment. I think we’ll see more challenges to the existing players coming from greenfield opportunities in the near future as customers have had enough and need alternatives. We hear that change will come from the edge, and it is already happening in the wide area network (WAN) space as software-defined WAN (SD-WAN) is quickly gaining steam.

Finally, risk is a factor. Nobody ever got fired for buying IBM, right? One would hear that from many customers in the past, but as the risk-averse folks in IT are starting to find out, there may actually be more risks in not doing something. While risk has been preventing customers from making moves in the past, this may be one of the biggest drivers for change in the future. Many of the customers I spoke to were actually more concerned by the risks of NOT doing something to change the current trajectory. As the business side of the house continues to drive for more agility (the current industry buzzword), we’re probably going to see IT viewing risk from a different perspective. More customers talked about the risk of not being able to move at the speed of the business as being the real driver for them (provided they can confidently ensure security and compliance first).

While some of the networking incumbents may have been looking at the market and thinking that they are dodging the bullet, our conversations at ONUG made it pretty clear that, while the bullet may not have the speed that some of the market challengers would like, it’s coming, and the trajectory is very clear. The networking of the future will be more open and more flexible. Maybe that is why we not only see the startups talking the open mantra, but the big guys like Cisco Systems and Juniper Networks are starting to preach that message as well. This is the beginning of some very interesting times if the ONUG meeting was any indication of where things are headed.