 Live from San Francisco, California, it's theCUBE at VMworld 2014, brought to you by VMware, Cisco, EMC, HP, and Nutanix. Now here are your hosts, John Furrier and Dave Vellante. Okay, welcome back everyone, live in San Francisco, this is theCUBE, I'm John Furrier, my co is Dave Vellante, our next guest is Chris Bull from America's CTO for VMware here, breaking it down. Chris, welcome to theCUBE. Thanks, John. Great to have you. Great to have you. Obviously I see you with your EVO Rails on the intro segment, we high as Dave is. EVO off the rails or on the rails, obviously you guys are laying out a vision and with some specifics around it. So with the dockerization, with open source, obviously apps are driving, and we pretty much called that, it's pretty obvious, we called it a while ago, but it is kind of on the stake in the ground, apps are driving everything. So the infrastructure has to be programmable, this is the number one thing that people are talking about in terms of the tech geeks. Infrastructure as code, programmable infrastructure on demand, whatever you want to call it, developers just want access resources, this is the cloud, this is also an on premise dynamic. So what's your take on that, EVO Rails, on the rails, off the rails, where's it at? It's 100% on the rails. This is our future unquestionably, and I'll tell you why. Look, IT professionals have much more important things to do than to do Monday in IT tasks, right? Storage provisioning, compute provisioning, network provisioning, security provisioning, and operational management are commodity IT tasks in the sense that they're non-differentiating, everybody has to do them extremely fast, and everyone in the world has to do them. I should not be trying to build a custom solution of what is a commodity IT service for every business in the world. I should have that delivered in software, I should have that delivered as software as a service, right, I should get an iOS-like experience to bring this software down and simply simplify those operations. Why? Because when I don't have to do these things, right, now my folks are free to do what's really important, to innovate, to drive business innovation, right, to research emerging security threats. That's what's important, and that's what EVO Rails is all about. So if I'm a person at a company, large company, and my job is provisioning infrastructure, and I'm really good at it, what would you advise me? I would advise you to start thinking beyond what you normally do, because that's not your value. Think about like electronics technicians 25 years ago, right? Those, we had folks that were really good, I was one of them, at finding transistors bad on a circuit board. If I was still trying to do that level of specialization today, I would be out of work, right? The industry's shifting, its agility is what's demanding these differences. So focus on what's going next, focus on integrating services of brands you trust, right? That's really where the value is going to be going forward. Chris, I got to ask you, first of all, it's music to our ears, we love the messaging and it hangs together beautifully, it really is relevant. Innovations about the next point of failure, you move to that next point of failure, you brought up security threats and mundane tasks. Automating away is an advancement trend, right? So when things are happening, automating mundane, repetitive tasks reduces things to automation, right? Talk about that piece and what is that next point of fear? When I say fear, I mean next pressure point could be security, so when you guys come in and say, hey, we're going to automate some mundane tasks, what is specifically those tasks that you guys are targeting initially and how do you guys sequence other things around the automation piece? Yeah, I think the easy ones, right, are things we've already been doing, right? Compute and storage, automation are big. Beyond that, if you think about where networking and security are going. Like if I'm going to take out my iPhone, right? If I'm going to call somebody right now, I don't know anybody's phone number, I might know two phone numbers today, right? What do I do? I tap on their name, but think about security, right? We're basing security on an arbitrary IP address, a number that can change, and people have to sit at CLIs and program all this in day after day after day after day. Why am I doing that? Security should be based on an application container, it should be object-based. People have done that because they've been doing it this way their whole lives, but give that up, right? Free yourself to really focus on what really matters. So let's talk about that, because Steve Harris was on yesterday, and again, great session, because he's just so candid. XCTO at VMware, as you know, is a great guy. He has a perimeter-less IT, which in a way, he's teasing out basically security, but the concept is not just around security. Could you talk about what that means? Because this kind of teases out kind of the direction that everyone's going. Yeah, I mean, it's complete overkill in traditional IT if you think about our security models. It's like saying, it's like, all of us have worked it for a company and somebody holds onto their job, right? It's like in New Jersey, if I go to pump gas, somebody will stop me and say, no, no, no, it's my job to pump your gas for you, right? I can't do it myself. They still do that in New Jersey? They still do that in New Jersey? What exit? What exit? Any exit. You can't pump your own gas, right? So there is definitely going to be some friction with people wanting to do certain things because they've always had that control point, but the industry's just shifting beyond that. I don't think there's a question. Now you gave a talk on Monday with Docker, better together was sort of the theme. Can you help us understand, sort of maybe sort of the premise of that discussion, but when VMware with containers and when best without? Yeah, you know, it's an interesting concept, right? Right now when you think about containers, it's mostly being used to drive third-platform applications, right? Stateless architecture, scale-out architectures, and that's where the focus has been. Now the challenge with the container approach right now, and I know Bengal will be on later, right, is that they do a tremendous job ensuring application and data portability, right? But the service stack that sits below that is still a problem. I've talked to multiple customers who have told me they love the portability aspects. However, to actually take the application and data and redeploy it somewhere else, to rebuild the service stack on new infrastructure with often new tools, it's two weeks optimistically to do it, right? It could take months depending on what kind of certifications I have to do. With VMware and having that consistent SDDC, that consistent API that I can run in public cloud, I can run in my private data center, I can run it in an outsource, or I can run in a colo, I can deploy these apps anywhere in the world, and I don't have to even worry about the infrastructure. Customers that are taking this approach are saying that they can see a full QA cycle on a redeployed application in less than a day. And the prerequisite there is some degree, Chris, of homogeneity, can you describe sort of that requirement? Yeah, you know, for us, right, what we're providing is there is a homogeneity aspect to it in the fact that you're opting into a VMware-based API set. You know, even some heavily outsourced customers I work with, they're dictating that they still want the outsourcer to run SDDC from VMware simply because it's going to give them flexibility if they want to leave that outsourcer's relationship. It's in the outsourcer's native interest to just say run on our custom stack because in the exit cost would be tremendous, right? But customers still want that assurance that, you know, if I don't like the relationship with one, I got thousands of other outsourcers globally that I can actually work with. Okay, so I don't mean this as a criticism, but I just want to understand it. So, right once run anywhere becomes right once run anywhere that has VMware stack installed. Is that correct? Am I understanding that? No, you're not. You're not, because right once run anywhere is a theoretical piece. Yeah, sure. If I have zero management, zero security, zero logging, right, zero capacity, zero disaster recovery, then it's technically true. Run anywhere means if I start something on Amazon and I move it to Google, I have to start all over again rethinking how I'm doing all those operational issues, right? They don't just magically appear again. So, that's that- Well, rethink and provision. That's correct. But right once run anywhere is essentially a fallacy is what you're saying. It's conceptually- It's correct. It's conceptually correct. In practice, it's not viable. Right, the operational stack always has to be there, right? We make that operational stack consistent in multi-provider. Other providers don't, right? If I'm a single service provider, right, even if I have a distributed architecture, I'm going to make it easier for you to go to my data centers, but I'm not going to make it easy for you to go to anywhere else because I'm giving you all the management and tooling. What we're saying is you can start in our data center, you can go to a partner's data center, you can redeploy it to your own data center, all the tooling and management stays the same and that's a huge cost savings. Right, so if I'm bad, I'm going to do a deal with you guys, I'll do a deal with Amazon, I'll do a deal with Google and that's how he's going to grow his tam. Right, right, yeah. I mean, and there's workloads where there's workloads where you don't care, right? So, there's workloads that I might just say let's deploy it to Amazon because I know it's going to live and die there and I'm perfectly fine with that. And what we're saying is, is that there's multiple models, right, there's other applications where I work with organizations where they have to get deployed globally, places where there isn't an Amazon data center, right, or it's a small office and that's what they need us for those flexibilities. So, there's going to be use cases where you'll use both in our opinion. So, what do you guys see as you have a solution of, let's just take a step back and talk about virtualizations roadmap. So, virtualization obviously been a great enabler. What does the EVO and the self-defined data center stacks that you guys are working on mean for virtualization for the folks out there who have been following it might not be in the, you know, inside baseball of what's happening? You know, what is this next wave of virtualization and what is it, is it abstraction away? Is it still multiple hypervisors, multi-cloud? What is the specific virtualization trend from your perspective? Yeah, I mean, the specific trend is the fact that, you know, we started on virtualization and basic server consolidation, right? It was all about the VM. The industry has moved, right? There's so much demand for agility today, right, that it's not just about provisioning the VM. And, you know, hats off to our competitors, that there's good hypervisors out there that can run a virtual machine. But that's just a fraction of the challenge today, right? What we're talking about with software-defined data centers, how do I instrument and virtualize storage, networks, and security, right? Things that everybody still needs. That's what's important and that's where the industry has gone. So when you can get all of that working collectively together, right? Now you have something really powerful. Our theme is, is when you think about infrastructure automation, it should be a feature. It shouldn't be a professional service. And that's what SCC is about. I can go and buy a competitor. That's your point about abstracting away the mundane task, right? Absolutely. And there's a huge IT services industry that has built, I mean, I've been in IT 22 years. My entire career is, let's encourage diversity and basically make everybody build their own custom solution and that fuels a multi-billion dollar IT services model. What we're saying is, customers are getting, are smart enough, they're figuring out that game and they're pushing back and they're saying, okay, yeah, they're giving all these pieces and parts, but what does it cost to build and maintain? And is that really the business model I want to be in? So let's just pretend I'm a CIO for a second, right? And we're in IT. We've been all in the IT world. We've seen the waves. Okay, Chris, you're on my staff. You're my chief IT guy, Chris. I really don't give a crap about the infrastructure. I got the apps coming in. I have, I'm investing a boatload of cash in all kinds of new apps with workloads that are in drive revenue. Certainly we've got some cost-cutting things going on with IT, but all I care about is I got to hire some developers. I want more apps. I want my customers being served. What does this mean for us? I just want it to work. So what do we do? That's exactly it. If you look at public cloud models today, the reason developers want to go to public cloud is because infrastructure is a checkbox with an API, right? It's code. That's what it is. And that's what we're working to provide here. So I get the entire infrastructure with a northbound API. Developers use the tools. If you stopped my session, you saw developers had no VMware touch, right? They were using all their native tools to actually do continuous integration and do updates and it was running on our stack in the back end. Our opinion is to automate that mundane stuff, yeah, we can do it really efficiently. And nobody is as concerned about, what are the interconnects there? Because it just has to work. It has to be the black box. But what's important is, I need choice of APIs and choice of interfaces northbound to hook into the developer tools. And that's why we're doing tremendous work around supporting OpenStack, something we've been working on for the last several years and we're going to continue to innovate there. So I got to ask you two weeks before we talk about the entry. You might have heard sitting around waiting, stateless applications is really a nice use case for Docker, developers love it. Stateful needs is also more of an enterprise feature. So if you, we love talking about the OS, right there. The OS is now the cloud, it's distributed. And I think Bill Fathers, I don't know, Ragu pointed that out yesterday in his keynote. So if that's the case, having a stateful component is key. How does Docker play well with that? And because VMware could add value there, right? So if I let the stateless and APIs develop and all that stuff gets more secure, whatever you want to call it, however that plays out, that's cool, that's working. I think the stateless, rest API, restful apps work. Great, right. Stateful is the issue. Docker, some people are like, hey, okay, we haven't seen that much here. So what do you guys, what's your take on stateless versus the stateful aspect? Yeah, I mean, when we have a stateful application, that's where you need to have services in the infrastructure that can provide the resiliency and the availability and the performance for that stateful application. So what we feel is even if you want to use Docker, if you need state, then we can provide those services for you, we can do VMotion under the hood, so you can still get the best of both worlds. You're going to get persistent data, you're going to get some of the hooks and integrations that Docker has as well. But to be specific, you guys see that as a real value ad for VMware on the state piece. Without question. Okay. Without question. But our real value ad is our massive ecosystem. It's the fact that we're the only vendor that can truly provide the entire infrastructure stack virtualized as a service northbound to a developer application, right? And beyond that, we're now providing through our project Fargo, right? A framework for extremely lightweight virtual machines that give you all of the benefits of the virtual machine plus all the benefits of the container as well at the same time. And we think that's going to be container Nirvana going forward. Talk more about Fargo. Can you explain what that is? Yeah, Fargo is going to, it's works in conceptually like a container architecture in that I can freeze and partition off child virtual machines just like I can freeze and partition off OS instances, right? There's a copy on right architecture with it. So now for each child, when it wants to write anything new or persistent, that's going to just stay in the child image. I can provision these children in less than one second. And we've already tested in scale to the hundreds here. So this is very significant in terms of massive provisioning very lightweight virtual machines. I'll have a shared memory state and I'll have a copy on right memory state per child. So this is really exciting technology. There's not a hypervisor vendor on the planet that can do anything remotely close to this. And this is not something new that VMware has been working on in the last few months. We've actually had this in development for more than four years. We saw the need coming, we've been working on this for some time and this is very mature technology. It's not a Hail Mary, you're throwing the Hail Mary here. Not at all, this isn't us, this isn't reactive at all because we've been working on this for more than four years. So the consistent theme I'm hearing is that relative to competitive hypervisors, like you said, hats off to them, but I don't think anybody's arguing that the maturity, the robustness, the ecosystem within VMware is substantially greater than anybody, and Microsoft included. Then there's open source, right? So basically you're laying out the roadmap for this industry. Just as Paul Moritz said, software mainframe. Basically caught us on a copy of that roadmap and a different software model. So the premise then is that you can stay ahead of that open source innovation for quite a period of time, right? I mean, from a business model, business plan standpoint. Is that? Well, it's a combination. Are you thinking about that correctly? We're innovating an open source too, right? Like look at the contributions we continue to make to Nova and Neutron and OpenStack, they're very significant. But let's be clear here too, right? There's some folks that are extremely religious about open source and I feel that the entire stack has to be open source and community driven. Let's call it 10% maybe of the world. 15% maybe. Yeah, and there's certain vendors that are very opinionated there. But here's the problem or the flaw in that argument. Our competitors, you look at cloud providers, right? Amazon uses open source components but they have a proprietary stack. And because of that, they can ship features in weeks. If we honestly think that we can sit on our hands and wait for a community to bless our features before we ship them, we're insane. I'm saying the premise is that you will stay ahead of that open source. Exactly, just like Amazon is, just like Microsoft is who has a proprietary stack. So we're straddling the fence, right? We want to be, we're being good community citizens, we're providing the open source interfaces that our customers want. And at the same time, we have to innovate through our stack. Now just one other point I want to make on that, you know, people say, well, what's in it for VMware? Why would we do that? Why not just circle the wagons, right? It might force people to stay in our API set. If we go open stack, it might be easier to switch out of VMware. And I'll tell you why. Our SDDC value proposition in the TCO savings we yield is so compelling, we're going to win that stack on merit. We don't need to worry about customers pulling out because we know they won't want to. So we're really putting our bet behind our technology. We're that confident that people are going to stay with us because they want to, not because they have to. So given that, if those assumptions are correct, that says the business model's intact for a long time. Relative to existential threats like open source. If you embrace it and you can stay ahead of the curve and you can balance that and earn the respect of the community, then you win. We are working hard to be a trusted partner of our customers and to do that, it means you have to truly listen to their needs, understand them and give it to them, right? When they need them. And that's our embracement of open source technology and open integration across our portfolio. So I got to ask you the hard question and what the crowd's challenging you on. I'll say that you guys have huge customer base. So you guys certainly do a great job there. So I really want to drill into that. But the questions that come in on Twitter, EVO rail, the more I understand VMware's EVO rail, the more complicated the server world has become. Kind of teasing out the counter argument that you're actually making it more complex on the server side. When simplicity and agile seem to be the goal, how do you respond to that question? The server world is complex. I met with the CIO about a month ago in Atlanta and he said, you know how many server and integration engineers I have? I said, how many? 65. He said, should I be employing 65 guys? I know that CIO. Okay, right? And I'm like, no, no, you know, you should not be in this business anymore, right? That's not where the value is. When you're paying people to do that, their focus is again on commodity IT tasks. It's not on driving business value and business differentiation, right? Which is where that money should be spent. So the future is in these appliance like technologies, whether that be through converged infrastructure, whether it be through hyperconverged, like our EVO appliances, because again, it takes out the mundane, it frees your staff to do the things that matter, to focus on critical applications. Senior Vice President of Phillips two weeks ago in theCUBE told us that 85% of the spend at Phillips, multi-billion dollar multinational company, 85% of the spend is non-differentiated. So everything's going to a consumption based model and all their vendors have to sign a deal that says, yes, we will do business that way. You guys actually signed it. Many didn't and they're not doing business with them anymore. They can't deliver the consumption model there. So okay, we got to wrap up. I'll see this impacts things like VMotion and other things you guys have going on. Certainly not, it's complicated under the hood, but if you guys can make it simpler, which your goal is, it's great. How do customers get started in this area? So let's just take EVO Rails or self-defined data center. What do you prescribe as the action item for CIOs and large enterprises that want to do the transformation? Yeah, John, I'm glad you asked that question because I get it a lot. And here's the thing, right? You look at the end state and you're like, wow, I love it. I love the value proposition of SCDC. I'm not sure where to get started. Think about, even if we consider EVO for like a 39 or a 395 type SLA, there's plenty of workloads that would run really well there. Think about development and test starting there. Think about your tier four backend IT reporting applications. Those are things that it can live on a lower SLA. You can get your feet wet there. Think about other workloads. Like if I have web scale applications and platform three applications, if the resiliency and performance is already built into the application platform, I don't have to run that on hardware that provides those same benefits. I can run that on a scale out architecture with locally attached storage. Now again, since we already have persistent storage in the platform, you can certainly run more workloads there, but most enterprises are conservative. So start there. Maybe you look at VDIs, another workload, and then you can continue to go up the stack from there. So pick something that you know has value to you, workload, go focus on that and then incrementally expand throughout. That's what you recommend. Without question. Okay, Chris Wolfe, the CTO for VMware for the Americas, former analyst, certainly in IT, 20 plus years. You've seen the changes. So I got to ask you, lay out the high level bumper sticker for VMworld this year. What is the game changing top line summary for customers and partners this year? What is the big bumper sticker if you had to put that on the car? We understand you and we understand the industry more than anyone else today. Okay, we are here inside theCUBE live in San Francisco, day three, wall-to-wall coverage, yesterday day in 41 interviews, setting records, VMworld's always back to back. We got no breaks till 12.40. Stay with us, we got more action coming. This is theCUBE live in San Francisco, VMworld 2014. We'll be right back.