x
NFV Strategies

How NFV Gets a Foot in the Door for SDN

A recent blog post by veteran industry analyst Tom Nolle points up where SDN is a tough sell to carriers: It's too ambitious, and the benefit is difficult to quantify.

Tom, who's a friend of Light Reading, starts with a reminder of my past as an English literature major. No, he doesn't ask if you want fries with that... he quotes the poet John Milton.

Then Tom gets into the meat of the thing:

When SDN got started, it quickly became identified with a kind of network-buyer "throw the bums out" vision where hordes of white-box revolutionaries were carried on the shoulders of the crowd as they trampled incumbents. In reality, there is no indication that SDN is really revolutionizing much at this point, and if you believe the latest Wall Street analysis of buyer sentiment, we're going to see an uptick in traditional network spending. Why is all of this happening?

Read Tom's whole post here: SDN is a Spectator Sport. (Also, see 3 Barriers to SDN Adoption.)

Tom then describes why SDN is having a tough time: "The barrier to SDN revolution is incrementalism. You can't readily value SDN by replacing one router at a time. You have to displace a bunch of stuff to create a value in networking, and to do that you need significant benefits -- new things."

Of course the big benefit of SDN, proponents say, is agility. You can configure networks faster, increase customer self-service and satisfaction, and innovate new kinds of service. But that kind of thing is difficult to describe in detail and quantify. And it's hard to explain exactly how SDN fits in, Tom says.

And as for innovation: The nature of innovation is you can't describe it in advance. The only way you know if something is innovative is in retrospect. Indeed, innovation often looks stupid when first proposed (and even stupidity is no reliable dowsing rod for finding innovation, because you know what else looks stupid when proposed? Stupid things.)

Reading Tom's post, it strikes me that NFV dodges the tough selling proposition of SDN precisely because it's less ambitious. NFV is about using software for network functions that currently require dedicated hardware appliances, such as firewalls and load balancers and such. That's easy to understand. And the benefits are easy to understand too -- the carrier doesn't have to roll a truck to install hardware on customer premises. (See Wind River Launches NFV 'App Store', Overture Adds Hardware to Its NFV Pitch , NFV-Enabled Ethernet for Generating New Revenues , and Cyan Debuts Planet Orchestrate to Manage Physical & Virtual Network Resources.)

There are other benefits too, but the CEO and CFO don't have to worry about those.

Like Tom, I'm reminded of the PC and Internet revolutions. A fundamental characteristic of those revolutions is they started small. PCs came into the office when accountants wanted to run spreadsheets without having to get the blessing of IT every time. The Internet at first just involved installing a web browser on employees' PCs, and putting up some company literature on the web, maybe sell a couple of products. Evangelists could build on small success and iterate until, soon enough, they'd upended the entire world economy.

But how do you get off to a small start with SDN? As Tom points out, it's hard.

NFV isn't so difficult. You can do a small NFV pilot, achieve a small but tangible benefit, and use that benefit to justify a bigger investment.

That's how revolutions are made.

So is SDN a loser? No. Operators at our recent Big Telecom Event said they need SDN to get the full benefits of NFV. With a single or a few virtualized functions, you gain some flexibility and economy locally, but SDN helps to deliver all the efficiencies and agility that NFV can, in theory, deliver across a broad network.

In that respect, the SDN and NFV revolution is like the PC and Internet revolutions that came before: You start out making small changes around the edges -- putting a few PCs on desktops to run spreadsheets, or putting your company's catalog and brochures on the Internet -- build on those benefits, until before long you're rebuilding the entire network infrastructure around the new technology. (See NFV 'Irrelevant' in Long Term, ONF Reckons.)

For more on some of the developments around NFV, see Exclusive: Leaked 'Inception' Document Fleshes Out Open-Source NFV Plans and NFV 'Inception' Meeting Highlights Tectonic Shift in Telecom.

— Mitch Wagner, Circle me on Google+ Follow me on TwitterVisit my LinkedIn profileFollow me on Facebook, West Coast Bureau Chief, Light Reading. Got a tip about SDN or NFV? Send it to [email protected]

sam masud 8/21/2014 | 12:45:59 PM
Re: Au contraire I'd say (optimistically) 3 year, realistically more like 5 years....talking about SDN in carrier networks.
kq4ym 8/21/2014 | 11:56:56 AM
Re: Au contraire I'm wondering what the time scale is for SDN to mature. Would it be the equivalent of going from oiffice PC to internet? It sure seems like the interval would be vastly less time, and if that's so, why not go into it now and not wait for a later maturity factor?
sam masud 7/18/2014 | 2:45:51 PM
Au contraire Mitch,

Many moons ago, the head of a very large IT department told me something that I've kept mind when it comes to new network technologies. Bascially, what he said was "that I can change all of the desktops any day, but don't mess with my network." This was at a time when there was a lot of noise about ISDN.

Although the above is not a very good analogy, you could say NFV is like PCs to SDN.

For all the hype, it will take a long time for SDN to mature...I only have to think back to how long it took to adapt a mature technology like Ethernet for carrier grade services.

 
TomNolle 7/18/2014 | 12:12:20 PM
NFV DOES Have it Easier I think you're right that NFV has it easier than SDN.  First, as you say, one of its value propositions--reduction in capex because of substitution of hosted functionality for expensive devices--is easy to understand and apply.  But second, NFV is about management and orchestration, which means that it has to solve those issues explicitly.  SDN is really about the same thing, but it occurs north of those "northbound APIs" where nobody seems to have much to say or offer.  If NFV gets MANO right, they could apply the lesson to SDN and be the real driver of SDN, not the ONF.
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE