 Welcome back to VMworld 2013. This is live coverage in San Francisco, California. We're in Moscone, South Lobby. This is theCUBE. This is our flagship program where we go out to the events and extract the signal from the noise. I'm John Furrier, the founder of SiliconANGLE. I'm joined by my co-host in this segment, Stu Miniman and our guest, Martin Casada. Your new title is Chief Architect of Networking at VMware. Also the founder of Nassir, obviously the major acquisition that kind of changed the game we were talking earlier. Welcome back to theCUBE. Great to have you again. Hey, it's great to be here, always. The Pat Kelsinger just said, Nassir was a shot hurt around the IT world and certainly virtualization has never been going away. It's only been getting better and more geeky under the hood. It's the engine that's powering the car of innovation, instrumentation with big data, applications, new stuff, it's upgrades coming, it's moving so fast. You guys really were an inflection point that enabled the software-defined enterprise, everything, the data center. Obviously networking's a lot of action being done. We've talked about that in the past. So I got to ask you, what's it like now? A year later, the game's changing, all the messaging, software-defined data center, hybrid cloud, end user computing is being enabled by a software construct that is based in foundation around the work that you did. Yeah, I mean, it's been fantastic. I think we need to do a journey like, you start with an idea and then you do kind of the engineering work and then you kind of build an organization around it and so forth. I mean, everything has its stage in the life cycle. When you're a startup, a lot of what you're doing is education and evangelism and kind of early sales and you end up with these kind of, I always call it the magic bean pitch, you're ready? So when you're a startup and you've got something new and it's a disruptive technology, the pitch to the customer is like, you know, I've got these magic beans and if you take these magic beans and you plant them in the backyard, they're going to grow into a bean stock and I promise at the top there's going to be gold and so forth, right? So it's this very evangelical way. Sell in a dream. Yeah, it really is, right? It's like kind of customer by customer and it's a lot of work because people don't understand this and it's new and it doesn't really exist yet. So there's a lot of faith and there's a lot of evangelism. What's been really wonderful about the last years, A, VMware is a phenomenal organization. It's got a huge customer base. You know, 500,000 some odd VMware customers and so what we do is we get to take a product and a technology, we get to have a sales force, we get to have conversations with customers which is very different. So instead of saying, I've got these magic beans and if you take these magic beans and plant them, we say, listen, you know, we already work together. You already use our bits. You know, you have problems in the networking. So let's work together to solve these problems. We can be very much customer led and we can work directly with them instead of like having to develop these new relationships. So Martin, it really seems almost it's gone from kind of the revolution discussion to, in the keynote you talked about a radical evolution. So is that a fair statement? Well, I think, I just think so. I think it's the nature of the association, right? Which is, you know, when you're in a startup you don't have like that relationship. You don't have that trust. And so, you know, so much of what you're doing is new and it's educational, it's building it. But from VMware, I mean, it's a very, very trusted company where we already have the bits. We already have the software. We already have the support infrastructure. We already have the sales channel. We already have the education and the content. So it's more about working with the customer to evolve it. It's a very different motion, but I think it's really necessary to take technologies and make the mainstream. And listen, this is all about making this stuff mainstream. All right, so Martin, last year when you were on theCUBE we feel like we're watching this together as we said, you know, you banged on the table said, you know, we're going to transform networking. Talked about how, you know, open source projects, you know, need to be, you know, more important than kind of the standards. You're doing a lot with open source still. You know, open stack of course is a huge engagement. But where are you guys putting your energy in open source? I saw some comments that open daylight didn't seem like a lot of energy, kind of watching it in open stack. Can you give us the update as to where you guys are putting your coding effort and your manpower? Yeah, sure. So we've actually like, on projects like OpenVswitch, which again is a project very near and dear to my heart, we've actually doubled the number of people working on it. And we work on things that are core to our mission, which is building the best virtual switch out there. But we also work on a number of other areas like open flow support that, I mean, right now we're very close to 1.1, 1.2, 1.3 open flow to support. We don't necessarily use all of them. I think it's very good for the industry. And I think it's very good for software to find networking in general. So OpenVswitch is definitely an area we put a lot of effort into. We've also, I think, more than doubled the amount of developers that are working on OpenStack is another effort that we work in. And at the same time, we still engage in trying to identify standards and API so that you can build the ecosystem around these things. Martin, we've always said, in fact we started when we interviewed you at OpenStack Summit, one of the things that we observed and then commented later is that the open source communities are now the standards bodies. And it's pretty clear that contributing code is now a way to get stuff ratified. Just push code, vote with your code. Right, that's right. And so I got to ask you a question. Why do you think OpenStack has been so successful? I mean, it's a framework, Pat Gelser calls it a framework. Huge initiative, people are rallying around it. Why is it being adopted so successfully from a mind share standpoint and also from an activity standpoint? No, that's a great question. So listen, I'm a huge fan of OpenStack. And I think that when we try and evaluate the industry around it and we try to think about what's going on, I think it's just as important to look at what's driving the adoption as what's being adopted itself. And I think in the case of OpenStack, there's this huge vacuum, which is a need for a composable framework where you can take best of great components or open source components and mix and match them to build your own orchestration and cloud management platform. I mean, I think in general, we want to move infrastructure from hardware into software. And in hardware, the interfaces are atoms, right? So you kind of know where the edge of a desk is or the edge of a cup is and you kind of fix these things together in a simple way. But when you move into software, if you want to mix and match and make composable systems, we need to have very well-defined interfaces. And I think the positioning that OpenStack took is listen, we're a framework of open interfaces that anybody can plug into. And this is the type of erector set we need. And it was a governance model, which is the kind of governance model we need to have industry participation. And so I think it's that need that people were really drawn to and this is why it's been so successful. And on the enterprise side, it's not too easy to build a cloud. I mean, I have a lot of the legacy and we talked to Sean Douglas earlier with Cerys-Mercy Bridge legacy and new apps. And CIOs, we talked to say, I like OpenStack because it gives me a warm and fuzzy, it gets a security blank, and we mentioned a binky for them and you know, nothing's the baby analogy. I'm getting killed today on analogies, but for CIO, they see Amazon disrupting. They see what it does with shadow IT and they see OpenStack as hope. But yeah, you got cloud stack out there, you got some other things, but OpenStack seems to resonate with the CIOs. What's your take on that customer perspective and why are they reacting that way? Yeah, sure. And I mean, I'm probably going to repeat a lot of what I just said, but I think it's actually worth reiterating, which is, with OpenStack is this promise of actually having composable software elements and this is what we need for infrastructure. If we're going to move into a realm where we're moving infrastructure into software, we want these Lego pieces that we can mix and match. And part of the core messaging of OpenStack from the very beginning was, we want open interfaces, we want composable interfaces, and we want broad industry participation. So we're going to do a consortium approach, not for example, a single company approach. And I think for anybody that's viewing these public clouds as the goal of where I want to be, and it's something that I want to control the way I get there by managing my own assets, this is a great message. I think it's a great platform to use. George Kadifat, H.P., who's now running the software group X Silver Lake executives on Oracle, he's just a great executive. We interviewed him and we asked him about OpenStack. We asked him a question, what are you most excited about? Just a general question, you know, a normal cube question. He goes, I'm excited about OpenStack. And I go, why? He goes, because no one vendor lock-in, no one vendor controls it. And he goes, and I go, he goes, do you believe in open source? He goes, yes, it's not about free, it's about freedom. And I think that is an interesting line. Coming from an executive at that level, that is about the freedom of choice. So functionality and lock-in aren't mutually exclusive. You can have some little bit of lock-in with the functions there, that's cool, but it was open, you still got choice. Yeah, yeah, yeah, that's right. And I think at the end of the day, the most important thing is as long as you know when you're taking something that's not standardized, as long as you know that, you can make a choice on whether you want to or not. So for me, the most important aspect of OpenStack, well, there's two very important aspects. The first one is community. No project exists without community and really projects are really the sum of the community around them. It's a great community, they've done great work. It's very diverse and it's really exciting. There's a lot of chaos and there's a lot of debates. It feels healthy, right? Community and the second one is the basic principles are around open interfaces and this is what we want. Martin, we could pivot to that discussion and talk about NSX, because we talked about VMware really becoming more of a networking player, but you're not going to do that along. You have partnerships, you're working to bring up an analogy, some have said that we're now entering the networking cold war because it's Cisco's vision of application-centric infrastructure and VMware's vision. We've had Brocade, Juniper, we're having Cumulus on today, all siding as to how they're integrating with NSX and Cisco is notably absent. So can you talk to us, what's the reality? Is your team working with Cisco and NCME and the partners, how does VMware's role in the networking growth versus the ecosystem? How does that all line up at your level? What do you say? I think it's really important to say that it's very important for us to partner with all of the networking harbor vendors. We don't do networking hardware, we will never do networking hardware. We want to integrate on open interfaces with all of them and we do. We have various partnerships and, in fact, it's undeniable that our partnership with Cisco is the deepest of any given the level of access that they've had to be sphere, the level of integration that we've had in things like VCE, right? And so we have varying announcements, varying degrees and the one that we did today was actually top of rack integration around a very specific product with NSX. But that's not meant to make a statement about who is preferred or who is not preferred. I mean, as we go through alliances and as we go through partnering, we want to partner with everybody, we're open to partnering with everybody and it's just, depending on logistics, we're on different points in these partnering tracks. Right, so can NSX be successful without Cisco at the table? Well, I think NSX in managing the virtual layer, absolutely because it is independent of a harbor vendor. It's just like, today VMware has V switches and has virtual ports and it has functionality in those. It has VShell Edge and these can be totally successful, independent of any sort of deep integration. I think that customers benefit through tied integration and I think that as customers demand these things, we do this type of work. One of the nuanced things that I saw in NSX that is getting lost a little bit is NSX can actually work across multiple hypervisors. Sure, yeah, of course. Open V switch was there, but when it's branded VMware, it's a little bit of a different discussion. So how is the work on different hypervisors different than when you were separate? I mean, I think we talked a year ago is working and integrating with ESX. Obviously it's a lot easier now, but on the other hand, has it changed the relationships with the other hypervisors? What do you see in KVM and HyperV these days? No, this is a great question. So just to be very, very clear, and I think Pat did a very good job when he did the keynote on this, which is NSX is a multi-hypervisor, multiple cloud management system product. I'm a huge believer, as I just talked about, OpenStack and horizontal integration. I always have been. It's fundamental in my system's roots, right? So multiple cloud management stacks, multiple hypervisors, right? And we've continued to do this work. We've continued to have customers that deploy in multiple types of environments. So that is from a positioning of the product that's from a support of the product perspective. But it's also worth looking at the logistics of the development cycle. When you control all the bits, you don't have to have consensus to make changes. And so we have a little bit faster innovation speeds because we control it, we do the release. And so, for example, when we do things in OpenVise, which especially at the kernel level, it's got to be upstreamed into the Linux kernel. And because we do everything open and everything that we want to upstream, then we have to make sure that we're following the correct process and architecting in a way that makes sense for the community. And so I think you'll see mildly different velocity speeds on the data path, but the goal is always going to be full feature compatibility on all platforms. The future of the data center is obviously the big messaging. So I'm going to find data center and obviously Pat's all in. I mean, he spent a billion dollars on your company, but the shift that happened, the shot heard around IT is common. It's interesting because, you know, we've been doing theCUBE here, it's our fourth VM world. So we kind of seen the evolution of the tech conversation, the business positioning, and this shift is really happening. And I got to ask you about the other elements, storage and compute. Because the compute thing came up in the future of IT, if compute is everywhere and is unused capacity, what does that look like? And that's kind of, I think, Andreessen Horowitz and Mark Andreessen's perspective is, you know, with virtualization, you can do some new things. Storage and compute are two areas that are being disrupted as well. What's your take on those? And what does networking, interfacing, if it's converged, they got to see some interplay. What's your take on that? Yeah, so let me answer the last question first, which is, so I think that there's a real temptation to conflate network virtualization and storage virtualization, which is if we virtualize the network, then we virtualizing the, you know, the substrate that goes to storage, and as a result, we actually get storage virtualization. So this is my personal opinion. I mean, of the opinion that's the wrong systematic approach. So what I like to say is you have to virtualize everything at its own layer where you have as much context in the semantics as possible, right? And so I believe that it's the job of the hypervisor to virtualize storage and it can connect to that storage system however it wants. And so from the perspective of a guest, the storage that it's seeing is coming directly from the hypervisor and the network is being virtualized through network virtualization subsystem. But this allows the hypervisor to connect through the storage using direct physical mechanisms without actually having to have virtual networking. So I think that two things actually work very well together depending how the system is architected. I also agree that there's a much larger disruption that's going on in virtualization of all infrastructure and we're still kind of coming to grips on what it means as these things operate within the same context, not just around how do they integrate, but like once you start giving this flexibility to workloads, how often are we going to be shooting ourselves in the foot, right? I mean storage always comes back, shared storage always comes back to these cache consistency problems and it's not clear to me that we've kind of handled this at an application perspective. So I think there's a lot of work still to do. Do you? Yeah, so looking forward, you know, we know it's still early days in this adoption. It was great to see some customers out there. Critique I'd say is those were really big customers. So you know, Citi, GE and the like, we'd like to see it go downmarked a little bit. What should people be watching for? What milestones are going to show that we're really getting broader adoption through the enterprise and maturing the stack as we go forward. Yeah, that's a great. So let me actually just address the point that you made, which is, I mean, when it comes to like, you know, press and it comes to marketing and it comes to positioning, you want to have the strongest position possible, right? And so, I mean, I think that the ability to, you know, bring these very large cross vertical companies on stage is a large testament. And I don't think that's a position on, these are the only guys that you can sell to. We've had almost an announcement, customer announcement a month that's going up market and down market. And so, you know, given, you know, whatever 17 minutes that we had on stage, right? There's only so many people we can cram up there. But I do agree that we're still in the early adopter phase. I think that we're still pre-casm and we're hitting the chasm right now. And so I think that, you know, it's not so much as the size, but where they are in the adoption cycle. And the way that we know it, I think it's going to be purely numbers, right? So I think we've already moved across verticals, right? I mean, even on stage, we had, you know, new service provider, we had kind of blue of the blue chips and we had a financial institution. And I think moving forward, it'll be nice when we see not only these early adopters and these different verticals, but we actually move into kind of more of the late adopters. Great, so I wonder if you can just talk a little bit about your team for a second. I'm sure there was a lot of pressure getting this out. Did you guys do a big celebration? And you know, how much pressure is there on you guys to deliver on that promise from, you know, a billion dollar acquisition? Well, first of all, the team. I work for the best team on the planet. Like seriously, I'm so lucky. I love the guys I work with, all of them. And I mean, even when I did research, honestly, I always prefer, I mean, I was always much more interested in the teams that I was working with and the topics that I was working on. It's the same, I mean, net reversalization is great and SDN is great and it's my passion. It's why I wake up in the morning, but like doing it with a team I have now has been phenomenal. But what's great, again, like I tried to describe before, we've kind of, the world has shifted. I mean, you know, from, you know, before the evangelical kind of hard sales pitch where there is a lot of pressure and you don't have the relationship and you have to build this all. You have to, you know, scrape it out with your own hands to now where these are conversations that we have with our customers and these are, and directly and we're helping them when we're working together. So you see progress every day. You feel very constructive. You're not doing these cliff hangers and so it's not really pressure. It's about execution now. So we're focused on execution. I think we do a great job working together to do it. Martin, I got to ask you about the show here. Obviously, BM World. What's your perspective going on? Share with the audience out there. You're in the trenches. You have a lot of passion. You wake up, you're awesome. And you have a lot of experience and knowledge. And you're a tech athlete, as we say. So I got to ask you what's, share with the folks what's happening at the show. What's the true signal out of the show from a tech perspective? And then what gets you excited? What trend and what tech gets you in the most excited right now? Okay, good. So, you know, my title's kind of changed within BM where I was chief architect and then sometimes CTO. Right now I'm kind of more in the CTO, but it turns out I'm actually a really bad CTO. I'm CTO of networking. Turns out I'm actually a really bad CTO. So I think that vision, especially in SDN, it's got its place and it's really important. But I think now we're at a spot where it's now time to operationalize and it's time to educate and it's time to have good proof points. And so for the NSX announcement, really it wasn't so much a vision announcement as it's here, people are using it. We've got a broad ecosystem, right? And then we've got a concrete roadmap going forward and here are the steps for you to engage the technology. So, VMworld in relationship to the NSX announcement, hopefully it's around kind of these very kind of concrete steps to move forward, right? I think VMworld more broadly and the messaging more broadly, there's a lot of focus on the software defined data center clearly and I think it's also in very much the same realm, which is just kind of very concrete ways to move forward but it's happening and customers are adopting it. What about some of the ecosystem partners? So they kind of shifting as well. I mean, also we hear software defined storage is big within the storage guys and then HP and others have huge announcements as well. What's going on in the ecosystem? What's under the hood that's driving? Is it flash? Is it some of the software approaches? What's under the hood? So I'm going to be honest. I'm the wrong guy to ask about storage because I don't do a lot with storage. I think ecosystem around networking it's very clear, right? I mean, coming with five or six partners, deep integration and like I mentioned in one of my talks and it's true, so one of the dirty secrets in the valley is anytime you want to do a press release and you really don't have anything to put on it, like you don't have customers or whatever, you just stick on a whole bunch of logos because everybody wants to be part of a press release. It's free press and you don't really have to do any work, right? It's true, right? So you're like, I have gone, man, I want to launch my company, I want to launch my product. Nobody's using it, probably doesn't exist yet. So you just go around the valley to all these companies and say, hey, do you want to put you to a part of a press release? And everybody says yes and they stick their logos on, right? This happens all the time. So as far as the ecosystem. We don't come with those. We don't come with those. So, but as far as the, you know, as far as the ecosystem that we're working with now, I mean, there's real deep integration work that's being demoed here today. So this is not just a bunch of people jumping on a bandwagon, it's actually showing real integration work. And I said, Martin, great to have you on theCUBE. I'll give you the final word. What's your plans for this year? What's your big ideas that you're going to be drilling down? It's going to be tactical. It's going to be exploring the ground. Bad CTO. I mean, I'm really focused on how do we bring this product into customer's hands? How do we operationalize it? And how do we move it forward? Well, congratulations. You guys have really certainly changed the industry. I mean, you're a data point, but it was a sea change that happened with software defined. You guys have been great success and bring it to reality seems to be the top job. Congratulations. Always have great to have you on theCUBE. Really appreciate it. Thanks so much. This is Silicon Angles theCUBE. We'll be right back with our next guest the short break. Oh, we're going to have Bill Fathers, the SVP here coming in talking, we're going to talk hybrid cloud. So we'll be right back into the short break.