 Live from Santa Clara, California. It's theCUBE, covering Open Networking Summit 2017. Brought to you by the Linux Foundation. Hey, welcome back everybody. Jeff Frick here with theCUBE. We are coming to the end of day two at Open Networking Summit. We just got here today. It's a great show. Everyone who's talking everything about software defined networking is here. And along with Scott Rainovich, we're joined by Dave Ward, one of the luminaries doing panels, doing keynotes. Here we are, in theCUBE. And here we are. Dave is the CTO of Engineering and Chief Architect at Cisco Systems. So Dave, great to see you as always. Great to see you guys. So what's the buzz of the show? You've been here for a couple of days. What, any surprises? No real big surprises, to be honest. You know, always there's some great announcements and great launches going on. But really what I'm finding surprising is that, you know, this is the sixth year of this conference. Can you believe that? So year six from where we started, and I may be the first person to say this, we, have you ever had anybody in theCUBE today talking about OpenFlow? You remember those days? No, we don't. No, nothing against OpenFlow, that's not my point. But think about how far we've gone. And so- Actually, yeah, Martin was talking about it. Of course he did. Of course he did. He's not going to let it go. But I love you, Martin. But really my point is, look how far we've come in six years. In six years ago we had a protocol, small community, one group working on this stuff, really working in standards. There was no open source associated with that at that time. Now look where we are. The, basically the place to do work is now in open source and come together as a community. So the buzz for me really is, holy shit. Like, this thing is real. There's a lot of people vesting a lot of money in time and really trying to work together to improve and build the ecosystem around networking, around network functions, what services have been delivered, and building a business off networking again. So networking is back, it's cool again. Right, right. And then there's this whole new thing coming down the pike in the form of 5G and IoT. That's just opening up a new opportunity, kind of redefine kind of what are these standards and how is this going to help kind of push things along? Well, it's kind of interesting. And so I'm just riffing for a second. So when you take a look at where we've come over the last several years and it was SDN controllers and configuring the network. Then it was virtualizing the network. There was a lot of talk yesterday and today about analytics and creating a reactive network. All of that has been built in those six years and come together in different open-source communities to build those pieces. SDN controllers, projects like Open Daylight, products like FD.io, projects like Panda, PNDA.io, and so that's the SDN virtualized network and data analytics piece. But when you get to 5G and IoT, one thing I'll be talking about tomorrow in my keynote is that there are big blocks missing in the industry. So let's dial it back historically. Remember when the HVAC contractor logged on to the network and that malware on that laptop stole 70 million credit cards? You remember that? Yes. Still haven't solved that problem yet. And so the reason why I'm bringing this up is what's missing, identity. So we had this notion that networks controlled by IT operators and they're going to go in to configure and provision that network. Well, we're now to the point where we need to link people and things to be able to drive what that intent is on the network. And whether it's buzzwords, which is real functionality by the way, of micro segmentation. HVAC contractor goes into a micro segment, can't get to the point of sale, can't steal the credit cards. Basic bread and butter stuff we want from the network. This is what SDN is supposed to deliver. Virtualized services like firewalls and others provide security. We'll just hold that for a second. But that linking of who the person is, what device they're on, where they are on campus, where they are in the world, et cetera, et cetera, time of day, whatever the case might be, are now the variables that need to go into the top of this system, into a policy engine that then drives that reactive network. So we've made a couple of great strides in six years, but to get to 5G and in particular to get to IoT, we have to have another couple of major blocks come into the industry to make that work well. And hopefully it's open source where that's going to go. And it's not just a standards body and not just open source, because we still need things to be manufactured and interoperable and the rest of it. So hopefully these things come together as we've seen the maturing of those two big groups. Right, that's great. I was going to say, it kind of begs the question, what is the interplay between standards, bodies, versa, or together with open source projects? Because before you didn't only have open source, the standards really set the regs. Now you've got these open source projects which have a main channel. They might start forking. There's all kinds of places that they can go. And how do the two kind of work together? Well, there's been a ton of effort and coming out of the SDN open source movement around model driven networking. And although it sounds kind of geeky, the main way of representing those models is through a representation called Yang. And the interesting thing about Yang is that's been not only adopted in SDN as the main object and way of representing the models being converted to networking equipment, computers, et cetera. But the IETF has taken that up and really driven a service approach through the IETF which is I want to deliver a VPN service. I want to deliver load engineering on the network versus what we did with SNMP or what the industry did which was I'm going to fully distribute this out to all the protocols and all the functions and everybody's going to write a maybe et cetera, et cetera. And we know how that turned out. So the creation of model driven networking, the standards bodies, picking this up, IETF, MEF, which is Metro Ethernet forum, broadband forum, BBF, these are all these organizations have now taken on that mantra that came out of open source SDN of model driven networking and are working towards creating those models. So that way we'll have a standardized way to program the network. But what's next is the telemetry coming out. Those objects need to be standardized. So that way, whether it's a Cisco device or somebody else's device is actually sending out the same data that can be collected and can be interpreted properly. Doesn't mean that it's a Mib. Doesn't mean that it's only going to go over one particular transport. I don't think anybody in the industry really cares whether it's JSON, Google RPC, Protobufs, Netconf or any of these pieces, they're all perfectly fine. They have different semantics associated with them. But nonetheless, those common objects and common data models have been what has been the key to keeping the industry working together, the common architectural philosophy and then the standards bodies have thankfully picked that up over the last couple of years. Yeah, we were talking earlier. I mean, you just threw out a bunch of alphabet soup there and I understand 80% of it. But it does raise the issue, we were talking about earlier about these standards development organizations and the IETF, the TM Forum, the MEF. Now we have open source. So we have the Linux Foundation. We have a lot of these different organizations and I think while you would know better than I as a CTO, people are becoming challenged by tracking and following all this stuff. Do you think we need some sort of consolidation of these standards or at least some more unification? We just saw ECOMP and OpenO merge so there seems to be some consolidation. What will we see going forward? What's going to help you as the CTO? There's no doubt, if there was consolidation that would be easier to track and easier place to develop. But in reality, Scott, it's 50 Shades of Yang. And the reason why I say that is each and every standards body has done their own specific function. Again, whether it's Metro Ethernet or it's broadband access or it's mobility, each one of those standards bodies is redefining themselves to be SDN capable. There's no doubt. If there was a one-stop shop, it'd be the most optimal way to get something done the fastest, but that's not the way the world works. So actually, I think we are going to see a continuous increase of more folks working on this, more foundations being built, et cetera, et cetera. Although what we have witnessed over the last couple of days in the last year is that the communities, the open source communities in particular are coming together and trying to integrate the pieces together versus just islands of cool technology that there's a few geeks interested in. You know, thankfully the operators and some enterprises have come in and said, I need this stuff to work and I need this stuff to work together. And that discipline is actually fundamentally new and different than the way other standards bodies worked or open source worked in the past. So I'd love to say that there'd be even more consolidation. There's frankly a bit of fatigue over, not saying it's whack-a-mole, but you have to really figure out and track where all this stuff is going on in the industry to really keep abreast and understand how wide and how deep it goes. It's interesting this trend lately where people are just donating, the project is just being absorbed into the Linux Foundation. So now there's at least kind of a consistency across all these various projects in terms of the way things are managed, the shows, kind of the communication and them helping kind of standardize a process to help those projects be more successful in their distribution and adoption in the company. Linux Foundation has done the industry a huge service. They understand governance. They've gone through a zillion different experiences of how to build communities, what works well when there's competing factions that need to come together and work, onboard marketing team, onboard legal team, able to build foundations as necessary or what's been experimented with over the last couple of years is, if you remember when we started to number these, you need to have a 503C, you need to have a foundation. There was frankly a high cost associated with these. Now open source is being contributed, there's no foundation and there's no cost. And so there's a whole continuum of things that the industry, the networking industry I should say is learning about how to build communities. And although this sounds cliche, you may launch a product, but you don't launch a community. You actually have to build it. And it's not all one company that's doing the donating or doing the working and that will produce, that'll create the longevity of that particular project. And that is what the Linux Foundation knows how to do well or at least catalyze people to come together to do that well. Now you mentioned like one of the big questions that always comes up with open sources. Well, how do we make money, right? Cause it's all free. It's like, you know, you know. Are we on Jerry Maguire? Yeah, so. Free like a puppy. It's still my favorite. Free like a puppy. Yeah, you guys still gotta change the newspaper, but so you were on a panel, I guess, which today there was a big discussion about the commercialization and how does, I mean, obviously Cisco has to stare at this big puppy in the room. If you will, you know, what's going to happen to our licensing model with all this open source? What came out of that discussion? What came out of the panel about the, how do you make money in this open source world? So a couple of things. One thing that was discussed was not only how to make money is which comes first, cost reduction, total cost of ownership or new service revenue. And really the outcome there and AT&T Comcast and Lightspeed Ventures was also on the panel with me. Needless to say, it's a combination of both. If you're coming in with a project and the project is, please spend this money so you can save this money, like we know how to do that math. We can add up the rows and columns and can understand whether or not money will be saved over time. But the new service revenue really, certainly in the enterprise space is really what's being discussed. And in particular, can I get these new services? I need these new security functions. I want to manage all my branches from the cloud or whatever the case might be. So new service revenue is depending on which use case, which technology, which layer, both of those two balance out and they both are required in the algorithm. Now, can people make money off of it? And the answer is, needless to say, Lightspeed Ventures colleague said, hey man, if there's a community and there's a technology, we can list off a zillion cases of where that community has turned into a true company that can provide value add and additional IP and move forward. Now, let's move this from just startups to big companies like Cisco or AT&T and Comcast. And not only do we all use open source in our projects, all those companies are contributing to open source and on Cisco's case, we're contributing to open source for a couple of key reasons. One is there are gaps in the industry, which we're limiting the industry. So let me give you an example. We open sourced a virtual switch router, which you might think, okay, it's Cisco, they're gonna do something in networking, but the reason why we open sourced it, and it's a piece that we actually use in our products was there was not a virtual switcher router that had the scale, performance or features that enabled the industry to utilize all the capabilities of the hardware underneath whether it's computer or networking or security. And so the industry literally would have stalled with a limited feature set versus being able to utilize decades of networking knowledge experience and things that are key and necessary encapsulations, features, filters, quality of service, et cetera, et cetera. There's a zillion of these pieces. And so there's a couple of different ways. How can somebody make money off of this really is the fundamental question. We contribute into open source communities and use that open source to build products as well. And we can do this across the video. We can do this in networking and we do this in NFV. We do this in orchestration in these pieces. And we also catalyze an ecosystem around these projects and then potentially around our portfolio as well. And so we continuously expand our ecosystem into startups that are using this technology, advancing the technology, enabling the industry to move faster and so on and trying to fundamentally create those business outcomes that our customers want. Just love that you just innately understand the value of an active community and it really comes through. So unfortunately the janitors have rolled in, the vacuums are going, the garbage cans are rolling. So before they unplug all of our gear, I want to give you the last word, Dave. What are some of your top priorities for 2017? So top priorities for 2017 really comes down to working towards filling the gaps I mentioned, identity and policy, but additionally number one, make sure that the automation orchestration policy around networking in a containerized stack is created. So we lived through a long era of hypervisors and what it was like to work with OpenStack and what it was like to work in an open source and have to invent all this technology. We learned a ton, but it doesn't exist in a containerized world. So for 2017, fill the big gaps in the industry and work towards orchestrating and automating networking, compute, storage and security in a containerized world. Pretty simple. I think that's the answer. It's going to say 42 is usually the answer, but I think that was it, Dave. Oh, I love 42, man. Brownie Lott, thanks, Dave. So he's Dave Ward, Scott Rainovich, I'm Jeff Frick, you're watching theCUBE from Open Networking Summit 2017. We'll see you tomorrow. Thanks for watching.