 Okay, welcome back to VMworld 2013. I'm John Furrier, the founder of SiliconANGLE. This is theCUBE, our flagship program. We go out to the events and extract a signal from the noise. I'm excited for the two guests here. We have Josh McKinty, co-founder and CTO of Piston Cloud and Joe Arnold, CEO of Swift Stack. And we've been discussing OpenStack for quite some time and obviously we've been following us since this foundation, prior to the rack space. And then this past year we had three days of live coverage with theCUBE at OpenStack Summit where we went up there, it was a full-documented, all developers, all signal, no noise, kick-ass, momentum. And of course we've been covering the space. But then we had the AWS debate. You guys were Joe, hosted, you were on a plane, everybody crowd chat, Randy Byas doing his thing. We missed you, Mark. We missed you. I'm sorry, yeah. So OpenStack is hot, you have Martín Casadausy on the religion drinking the Kool-Aid. VMware saying, hey, we love OpenStack. It's a market expansion opportunity. It's kind of out there. So guys, let's talk about OpenStack in context to VMware. Big buzz at OpenStack, Martín essentially laid it out. It resonates with people why they want it more than what's actually happening, which is still impressive. OpenStack resonates with enterprises and the communities behind it. So that's positive. So how does that translate into VMware, Joe here, VMware? This is a great question. I think Pat put this really well yesterday when he was talking about legacy workloads and next-gen workloads. And the comparisons to mainframes are perhaps harsh, but fair. VMware and Microsoft compete for the mainframe workloads in today's data centers. The equivalent of the mainframe, which is traditional client server applications. And the perception. Which is now cloud, right? The perception these days is that every market is going to be a two-horse race and the two-horse race for next-gen cloud applications is between OpenStack and Amazon. VMware is not going to be a third horse. They have to be part of the OpenStack team and I think that's the transition we've seen them on in the last year. And I think they've voted with that and you're seeing the positioning and posture is very positive right now. People are smiling and palm pressing, getting behind each other. But some are speculating that that might shake out as people see the land grab opportunity possibly. Well, look, what we're seeing is people are just using the best tool for the job. And you're certainly seeing your deployments co-mingling with VMware deployments. We're seeing exactly the same thing. So what we do, what we do, that Swiss stack is more focused on the storage angle and that can co-mingle with all sorts of different environments. And as the infrastructure that is getting built out gets bigger and bigger, they're not applying multi-tools or general purpose things anymore. They're finding the right tool for the job in the data center. I think what you guys are doing is awesome and I think one of the validations I'll share with you we had an HP event last month. HP big data event held the top customers all and they were big names. They weren't slackers and they weren't just there as plants. They were like big name companies, you know, Facebook and a bunch of others, Twitter, you name it. All the big companies, Inga, you name it. And the feedback from George Kedifa who runs HP software was, I asked him, what is the most exciting thing that gets you jazzed up? And I said, pick your answer, you know, tee it up. He said open stack. Of all the choices he's open stack, I said why? Because open stack, it's not about the free software, it's about freedom, right? So that's, to me, really speaks volumes. And it's interesting, people want agility and service catalogs, but they need a foundation. So I want to ask you guys the question, what foundation elements need to be in place? And you see vCloud kind of telegraphing their position with disaster recovery. It's kind of like, I don't think anyone's standing up saying I'll volunteer and do disaster recovery for open stack or are they, is they going to go after the easy foundational things? And what does a foundation look like for the proprietary vendors? I mean, this is an interesting question for both of us because I think Swift Stack and Piston have both focused on curated fundamentals. Which are those foundation pieces. You know what we do? We do HA, we do an update service. We do these really boring and very necessary pieces. We do security. Boring and table stakes for the enterprise. Table stakes, absolutely. You need, what's fun about this collection of fundamentals is we have joined customers now with basically every ecosystem player. So we have customers with Nasira, which I think you saw our logo during the keynote. We also have customers with other network vendors. And Ditto on the storage side. And so that really is that freedom that folks are looking for is to be able to mix and match a set of vendors in their data center. The software defined data center label is a really powerful concept because we're taking the silos down around storage and compute networking, but we're not limiting the scope of the vendors and saying okay well because it's all one set of APIs that doesn't mean you have to buy from one individual. Joe, talk about this from your perspective because you mentioned tools for the job. Martin talks about an erector set. And we all as kids play with erector sets, we're dating ourselves. I don't know if they're still around, but that's kind of the mindset. Hey, build your own, but here's some base frameworks and more solutions. It's all about the private cloud and our boring features, LDAP integration, package-based install, we're collectively I think trying to make it so that really any operator can get up and running with this infrastructure. And so, I know this kind of changes the software. So Dave and I were, this is the first time we're talking about on theCUBE, but I'll throw it out there because it's a half-baked kind of thesis, but we were speculating open sources obviously the future. It's ratifying things in the community. That's the new standards bodies we said that, but the old model of downloading Linux, the new generation of guys love Amazon because the stack's all being updated for them. So that kind of brings back the appliance. So if it's a tool for the job, do the customers care that there's another appliance that kicks ass on, say LDAP integration? Well, it's all about building out private cloud infrastructure and what does that private cloud infrastructure look like? And in order to get a footprint into the private enterprise, you need to have those integrations like with LDAP and Active Directory and ease of deployment and management and operations. And what we're seeing from a storage perspective is that people are bringing their storage workloads in-house and that co-exists with a core of compute infrastructure which actually exists with many different vendors. And then from a bursting perspective to go out into the public cloud, yeah, you can do that. It's like this analogy between water and electricity that we have. It's data is like building those dams and aqueducts. You guys see Pat at Gelsinger's interview on theCUBE this morning? No, I didn't. Same analogy? No, no, no. Well, you should watch that interview. But I want to ask you the definition of private cloud because obviously VMware's betting on hybrid cloud, which I'm a big fan of. Don't get me wrong. I did kind of make a reference to it calling it a halfway house and Pat kind of punched me in the gut a few times on that. But I meant it in a very, you know, way of it's a destination stop. It's a way station. Some think, oh, that was my perspective, Pat. Pat disagrees, he's betting all in on hybrid. Sure, look at the water analogy. What's the definition? Is it different? What is private cloud infrastructure? Is it? Look at the water analogy because everyone has a hot water tank in their house, right? The reason is, yeah, you don't want your own well necessarily. That's not efficient. You can rely on other people to get the water out of the ground. But keeping the water hot, that's something best done local. That really is the definition of what will be the long-term private cloud solution. Private data belongs in private places. Typically in a building you control if you really care about it. And data locality does matter for performance. So if you're talking about big data and this keeps coming up as a use case, big data really latency matters. Putting it in someone else's data center is probably not the best solution. We have customers that had a huge footprint in Amazon and it was incredibly expensive for them to deploy, maintain and operate. And by able to shrink that down and doing private data center, they've been able to optimize that from a price and a performance perspective to do exactly what they needed. Yeah, we had George Sussman from io.com which is a new hot company. Well, they've been around for a while. It was a self-funded start up, great guy. And he said, hey, software defined data center is great. But what about the data center piece of it? Software, obviously everyone's getting behind the software piece, which we love. But they still have that data center, the hot water tanks, if you will. So that is an interesting film. That's power and cooling. That's footprint, that's sensors on devices. That's moving workloads around. That's still who the action is. I don't want to move this conversation all the way out into this converged infrastructure domain and talk about the ODMs and Quanta and Open Compute. But that is this other wave that's coming through the data center right now. And I think if you look at IBM and HP and Dell and how they're positioning around OpenStack, they're driving up the stack as fast as they can. The same thing that VMware has been doing, driving towards those management tools. And the reason is because the physical hardware is both converging, the storage and compute networking hardware, and it's losing its margins. When you're competing directly with an ODM who are happy with a 10% margin, if your business relies on 30 to 90% margin, that's going away really quick. We have our biggest customers, when they said, hey, we'd like to work with your hardware vendor to make sure we're compatible. They said, which of the three? Yeah, and the same with storage and all compute. So all pieces of converging infrastructure. So OpenStack gives that freedom, right? So that's back to the freedom, not just free. Nothing's ever free in life, as we know. But OpenSource is great, but let's talk about the freedom piece. What specific things do you guys point to right now that is the best about OpenStack and from a tech perspective under that allows for freedom for construction of public cloud, private cloud, and hybrid clouds for our customers? Go for it. So I think at the core, all of us are building at the foundation of what we do. The engines behind our product is based on OpenSource. And that gives our customers and the whole ecosystem to participate, to contribute to it, to move it forward as a shared group. And so I think that truly represents freedom because there is options. They're not going to get locked in. They don't want to get Microsoft again by being locked into a single platform. So having, going with OpenStack gives them that choice. I would double down on that comment and say one of our biggest customer deployments right now has, and in fact 80% of our large deployments are using Cloud Foundry on top of OpenStack. And they're looking at this freedom from the past layer as well as the infrastructure layer. They have two or three different SDN solutions. They have a couple of different hypervisors and they're saying, you know what? The dashboard that we're relying on, the orchestration tools and the past layer is unified. We can rely on that and we're going to experiment under the hood knowing that we can double down on any one vendor later on. And we had two guys on earlier that I really like to talk to. One is Sean Douglas who has a lot of experience at EMC and CTO office and EMC Ventures. So he spent the time walking the landscape and also involved in the New Sierra deal and Martin. Right, so we always talk about the hypervisor. Now you got Sean, his loving life right now at Service Mesh and Martin's obviously executing out on the billion dollar bet that VMware is making. But they all talk about the virtualization around the hypervisor. So what is the current state of the hypervisor debate right now in terms of how it should deal with multiple hypervisors and storage and other elements? I love to keep this in real time. So there's an email thread this morning on the OpenStack mailing list proposing to add support for Docker to OpenStack. Docker for those who aren't familiar with this new sort of container based pass. People are starting to say, you know, what's really the difference between a VM or a container or bare metal provisioning as far as the user is concerned? Remember this entire cloud movement is being driven by DevOps. The developers don't care. What they want is some logical container for their application components. And that might be a process. It might be a container. It might be a VM. It might be a full machine. So OpenStack more and more is allowing folks to deal with all of those resources in exactly the same way. That's super exciting. Yeah, it scrunches down the importance of what that hypervisor is. And the importance is in the applications and the services that those applications can consume. Whether that's compute resources or database resources or storage resources. So the diversity of containers is a better approach than trying to make someone bite off more than they can chew. We can abstract everything below what that containerization looks like. What that platform looks like. So down here we can swap things out. We can try new things. So we can get the best tool for whatever application that we're trying to build. And so what does this do these kinds of debates? I mean, one thing that scares me about standards bodies or communities is that you have basically gridlock from people trying to stall the process of innovation by throwing more arguments onto the fire. OpenStack has done pretty well on the governance side. Are you starting, is it still solid? Are you guys happy with that? Do you see more hay makers coming in? I am more and more happy with it because proof is in the pudding. We've got customers who are happy with it, right? We continue to deploy this every week. We're like standing up another large environment. And that to me is the final proof point to say we're not a standards body. We've never been a standards body and we've been very careful about not accidentally turning into a standards body. What we have is rough consensus and working code. You know, it's interesting. We love doing the cubes. We get to hear from everyone, especially the smart guys like you guys. But last night we were talking to the VCs who are, they have a lot. They have big checks, they write checks and all the parties were last night. So the conversation was like, okay, well what's the future of compute, the containers? And they're trying to make, they're trying to bet on a market. It's a moving train at this point. So it's really hard. So what are you seeing on the startup scene? Because you guys are out building companies. So what is, what's your take on the startup scene? Are the bets already made? Are there new bets? I mean, do you see the young guns coming in? The DevOps culture is now a mindset. It's now the Kool-Aid, everyone's drinking. Look, I mean, we're, I think, I think Joseph and I are looking at each other because we're just in there trying to duke it out, trying to solve customer problems. And yeah, our respective boards are probably outsourcing and looking at deals. But that's not something that I guess we're plugged in at. We're actually on customer sites doing deployments, getting infrastructure stood up and just doing the best thing we can there. The beautiful thing is the market is so wide open right now and there's no lock-in anymore. It's like Dave and I were talking about the car analogy. We're like, yeah, it's kind of like horse and carriage has gone, but now you got cars, but there's so many different car types you can make. You can make sports cars, you can make vans, whatever you want, you can build the engine of virtualization and conversion infrastructure with softwares, interesting. Because now it's like, everyone's talking about the engine here, right? You know, not necessarily the dials and AC or power windows. And that's a big deal. That could be the applications, right? So again, this comes back down to the fan base, right? So with that analogy, is the engine almost ready? What's the engine look like? You guys are always considered in that department. What is the engine? The engine is more than ready. I mean, the cars have been rolling off a lot for a year, right? So the race now is to make sure that they handle well and that folks aren't driving over the speed limit. So if you think about, okay, what were the first Model T's like? No seat belts. So yeah, they're ready. People are driving them. Now we spend a lot of time. I'm putting graces as a starter. It's got to turn. I mean, the screen is just making sure of, it's about the apps. It's like that economy thing. It's about the apps stupid, right? That's what people are doing this for. Follow the apps is what we said. Follow the apps, follow the money. That's the kind of the paradigm. Absolutely. Okay, we got a break guys. Joe and appreciate it. Josh, thanks so much for coming on theCUBE. We'll be following you. We're not going to go to the OpenStack Summit in Hong Kong, a little bit through the long haul for theCUBE. Unless someone ponies up some underwriting, maybe we'll consider it, but we'd love to go there. We'll certainly be at the next one in Portland. Congratulations. It's great to have you on the chat with Randy recently. So good luck and congratulations. This is theCUBE at VMworld. These are the guys making it happen, solving customer problems. And that's the proof in the pudding. Voting with code, solving customer problems is the start-up opportunity for all you entrepreneurs out there. We'll be right back after this short break. Thanks, John.