Ideas coalesce all the time in every vertical. You don’t really notice it until you wake up one day and suddenly everything around you looks identical. Wireless becoming the new access layer. Flash storage taking hold of the high end performance crown. And in networking we have the dominance of all things software defined. One recent development has coming along much faster than anyone could have predicted: Software Defined Wide Area Networking (SD-WAN).
Automatic For The People
SD-WAN is a force in modern networking because people want simplicity. While Ivan does a great job of decoupling marketing from reality, people still believe that SD-WAN is the silver bullet that will fix all of their WAN woes. Even during the original discussions of SD-WAN technology at conferences like ONUG, the overriding idea wasn’t around tying sites together or driving down costs to the point of feasibility. It was all about making life easier.
How does SD-WAN manage to accomplish this? It’s all black box networking. Just like the fuel injector in your car. There’s no crying about interoperability or standards-based protocols. You just plug things in and it all works, even if you can’t exactly plug one vendor solution into a competitor. Lock in wins again.
The ideas behind SD-WAN aren’t exactly new. Cisco talked about SD-WAN quite a bit at Networking Field Day 10. Here’s Jeff Reed on it:
The rest of the two hour session details how Cisco is using their Intelligent WAN (IWAN) product to drive SD-WAN. The names of the components all sound very familiar to networkers: DMVPN, NBAR, PfR, and so on. That’s because SD-WAN uses a lot of tried-and-true techniques to tie the concept together. There’s nothing earth-shattering about SD-WAN under the hood. In fact, a fair number of people that work at the “pioneering” SD-WAN startups all seem to have their roots in one or more traditional networking companies.
Fables of Reconstruction
Look at the other presenters at Networking Field Day 10. Two of them announced SD-WAN solutions even though they aren’t really known for expertise in SD-WAN. One of them wasn’t even known as a branch office acceleration solution. So why the SD-WAN land rush all of the sudden? What’s behind the need to have a solution?
You probably wouldn’t be surprised to learn that a lot of investors are backing expansion into SD-WAN technologies. It’s a hot property. But why? As above, customers aren’t interested in the technical wizardry that goes into SD-WAN. They aren’t clamoring for it to supplant their current WAN solution and offer a Rosetta Stone of inter-vendor WAN cooperation. What’s behind the push?
It probably goes something like this:
- Technologist needs to implement WAN architecture. Is dismayed that things are so difficult.
- Technologist starts searching for solutions about WAN. They probably start asking friends about it.
- Analyst firm hears that technologists are asking about WAN solutions. Releases a questionnaire asking which technologies you’d like to learn more about.
- Responses to questionnaires are loaded into a graph or report that people buy because they don’t know who to talk to.
- Companies realize customers want WAN solutions. They break their necks to offer those solutions to keep up with demand.
- Investors see companies beginning to offer WAN solutions and think there’s a huge untapped market. They start funding anyone that mentions WAN in a meeting.
By the way, you can replace “WAN” with any technology above and it still works.
Thanks to customers needing a solution for something they can’t configure easily they are going to be inundated with SD-WAN options by the time they turn around. And the biggest concern no long becomes “Who has the easiest solution?” but instead, “Who is still going to be here in six months?”
Collapse Into Now
The reckoning is coming in the SD-WAN market. If a company doesn’t already have an SD-WAN solution in development or if their solution won’t see daylight for another nine months, they are going to exercise the second “B” of innovation and buy it. And they have a lot of prime targets to choose from.
Investors get cagey without an exit strategy. How are they going to win at this game? They either have to get paid with an IPO, with a later round of funding, or by having someone buy out the investment. If an investor thinks they can get their money back (plus a bit of interest) by having this little startup bought by a traditional networking vendor you can better believe they will be advising the startup to sell.
The customers are the real losers in the case of a buyout, or worse a bankruptcy. Those highly proprietary solutions become dead weight if there isn’t any support for them any longer. Black box networking falls apart when the little magical creatures inside the box go away. Which means customers will be skittish of supporting a solution that is likely to go away any time soon.
Who will you support? An established vendor slow to roll out a solution? Or an up-and-coming company with new ideas but at risk of being snapped up by a big bank account?
I loved seeing all the SD-WAN discussion at Networking Field Day 10. SD-WAN is no longer magic sauce that aggregates DSL and MPLS circuits with encryption. Nuage Networks showed off deploying Docker apps to remote sites. Riverbed talked about using their WAN optimization experience to deploy SaaS solutions through SD-WAN.
We’ve heard from SD-WAN companies in the past at Networking Field Day. It’s interesting to hear the comparisons between the upstarts and the old geezers. It’s clear there is a ton of money that is being invested in SD-WAN. The trick is to find out your needs and pick the best solution for you. Otherwise you may find yourself losing your SD-WAN religion.
This space is a little frustrating right now to be honest. I am in the middle of designing a small, but complicated WAN deployment that perfectly suits the SD-WAN solutions, I just can’t quite bring myself to recommend any one of them at the moment. Who knows if Viptela and Cloudgenix will be around in a couple of years, and under what form. Riverbed doesn’t have anything shipping yet, neither does Cisco. Nuage is a little overkill for what I need… Maybe I’ll go for Silver-Peak… Or just go down the DMVPN/FlexVPN//PfR rat-hole and hope that Cisco actually delivers in the end I can migrate then…
Not an easy one. If I had 100 sites to do, this would be easy.
A shameless plug and a disclosure here. I work for Viptela. Viptela’s technology is grounded in the principles of networking that kept the world of communications going for the past 3 decades. Yet, it provides an overall level of innovation that is currently unrivaled. Trust me, I had been around the block, so I am know what I am talking about. Reachability, security, application policies, all at scale and with operational ease and visibility.
Small environment? No problem! From a few single sites to hundreds and thousands, you will not know the difference. Same solution and architecture ubiquitously applies to all.
Don’t want to redesign it all at once? No problem! Our solution is fully routed architecture that can fit in the most sophisticated environments out there today. We know, we already deployed so many of them 🙂
Worried about investment protection? No problem! Our solution seamlessly leverages your existing investment in routing gear. We have many design options.
I wouldn’t worry about Viptela not being around. Technology we pioneered, partnerships we built and customer traction we have is a solid proof you invest wisely.
If that sounds good and you want to learn more, drop me a note at firstname.lastname@example.org.
David, I am already in contact with Viptela folks. Thanks,
Yo yo yo, where exactly is the snark fella?
Pingback: This WAN Is Your WAN, This WAN Is My WAN - Tech Field Day