Recently transplanted from Cisco, Meyer talks about SDN's audiophile potential and the ad-hoc implementations of early adopters

Craig Matsumoto, Editor-in-Chief, Light Reading

December 21, 2012

6 Min Read
Q&A: David Meyer, Brocade's New SDN Expert

One fun aspect of software-defined networking (SDN) is on the personnel side. It's created a rush to grab experts in switching and routing protocols.

Cisco Systems Inc. plucked David Ward from Juniper Networks Inc., and in November, Brocade Communications Systems Inc. got David Meyer out of Cisco. (See Cisco SDN Expert Leaves for Brocade.)

Light Reading got a chance to talk with Meyer by phone this week, to get his perspective on the amorphous identities of SDN and find out where he expects to help steer Brocade.


Light Reading: So, how's the job? You've been there a couple of months now.

Meyer: That's right, I've been here about nine weeks. I've been traveling every week. I live in Oregon, so just to be here [at Brocade, physically] I'm traveling. In Europe, I've been at industry events I was already committed to, and I've been out to see customers. Asia, too. I've been seeing customers, getting a feeling for what's going on on the ground.

Light Reading: Why'd you move from Cisco?

Meyer: I thought there was a really great opportunity here at Brocade. I think we have a great opportunity to build our business, especially in the area of SDN. This is going to be about software, but we also have a good substrate for this both in hardware and software.

And great people. Not that Cisco wasn't great; Cisco's a great company. The biggest thing is difference of scale. If you look at market cap, it's about two orders of magnitude. It's kind of a rule of thumb; everything's two orders of magnitude smaller -- headcount, etc. -- and that level of scale is really noticeable, in ways that are positive and other ways that are not.

Light Reading: So, what's your definition of SDN?

Meyer: My view of it is quite a bit wider and deeper than the sort of SDN I've been calling OpenFlow SDN, just to get a name for it. OpenFlow SDN is one piece of the bigger programmability picture, and that picture is all the stuff that's going on in the IETF, the IEEE, the new ATIS exploration [see ATIS Weighs In on SDN Standards] -- and in the ITU, for that matter.

I'll give you an example: You might decide one way to do this version of SDN is to peel off traditional pieces of the control plane and do them in a centralized way off of the new architecture -- something like PCE [the IETF's Path Computation Element model]. That's a piece of SDN that isn't necessarily OpenFlow. The network is getting programmed by the PCE server, and the PCE server can be hooked up to anything it wants, including an optical system. Above that, there's an entire universe of provisioning, orchestration layers, CloudStack -- all that stuff.

So, my vision of programmability is an ecosystem. Brocade's playing at the bottom layer of this ecosystem, where you have forwarding layers and control planes.

Light Reading: Most of the SDN talk focuses on openness and the need for industry standards, but -- how "open" is this really going to be?

Meyer: It's created an ecosystem where everyting is, I would say, a little bit de-siloed, where everybody has to play with everybody else. If you go to a place like VMworld, you'll see VMware Inc. would announce something that's made up of five or six different players.

What I've seen recently is that some people -- they're like audiophiles. They want to get the best, and they want it to work together. So, there's a lot to choose from at the higher layers of this ecosystem. It seems like this ecosystem is growing at a high rate. Except for probably the non-commodity switch level.

Light Reading: Well, even there -- it depends what you mean by "commodity," but the Broadcom Corp.-based switches, the companies like Arista Networks Inc. -- they'd been built up quite a bit.

Meyer: Sure. No matter what you do in software, eventually somebody's going to have to switch packets. That part's not going to go away.

To Page 2 Light Reading: So, how do service providers view SDN? What are they saying about it?

Meyer: They want two things.

They want to know how to better optimize their networks. Even the most advanced carriers are using decades-old technology. They need ways to look into the network and optimize what they see. And the other thing is: They want to monetize. They're just getting killed by the over-the-top players, and the reason the over-the-top players gained agility and speed is because they don't interact with the network.

Every carrier or service provider you talk to has the same problem. Everybody wants "faster," of course, but once you get past that, it's all about how you want to optimize and monetize.

In order to optimize, you have to look across all layers of the network, and that's not possible right now. And to fight the over-the-top players, they could monetize data in a couple of ways. If you provide ALTO [an IETF draft protocol] and some kind of network positioning service in your network, then the network could tell you where the closest data center is. That could provide value if there are latency requirements. The other thing is privacy. We've seen where service providers have been able to extract data, anonymize it, and then sell it.

There's a new life cycle for the carrier network that's emerging, where you data-mine the network, you send the data to some kind of analytics farm, and you program the state back into the network, and you seed this in real time. It's really the only way the traditional carriers are going to be able to do the opimization/monetization stuff they want to do.

Light Reading: Hey, that cycle -- I've seen that. That's that triangle diagram David Ward shows. (See Cisco Links SDN & Policy.)

Meyer: It's a pretty obvious thing, and you have to be able to do it. Everybody's showing that picture.

Light Reading: How easily will service providers be able to execute these things?

Meyer: This optimization and monetization thing -- they're not going to be able to do it soon enough. Any one of these carriers is in a fight for their lives against someone like Google. They need this now, and they don't have standardized ways to do it, so you've got all these ad hoc ways out in the field. I see this on all continents, everywhere.

People are writing the higher-level pieces and using API [application programming interface] gateways -- Apigee Corp. or something like that. They build these gateways where maybe you have an API, and they can reformat it into something like, say, Netconf.

So, that stuff is all happening right now. I talked to a few large service providers in Europe that have a large part of this ecosystem built, and they built their own OSS system on top of it. By having standardized APIs, it imposed discipline on the OSS/BSS writers in a way that forced them to have more portable code, cleaner code. It's happening now, and it's accelerating.

— Craig Matsumoto, Managing Editor, Light Reading

About the Author(s)

Craig Matsumoto

Editor-in-Chief, Light Reading

Yes, THAT Craig Matsumoto – who used to be at Light Reading from 2002 until 2013 and then went away and did other stuff and now HE'S BACK! As Editor-in-Chief. Go Craig!!

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like