 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 here, live in San Francisco, California. This is theCUBE, our flagship program. We go out to the events and extract the signal from the noise. I'm John Furrier. Founder of SiliconANGLE, my co-host, Dave Vellante, co-founder of Wikibon.org. Our next guest has been Gallup, CEO of Docker, star of the show, star of all the shows we've been to the past year. You can talk DevOps, application development without kicking in and talking about containers. Containers without compromise, Pat Gelsinger set on the stage. Ben, welcome to theCUBE. Hey, thank you very much for having me. Great to have you on. I've known you've been following your career for some time, going back to the Verisign days, and you're an industry vet who's been around the block. You've seen some innovation cycles come and go, certainly riding high right now on cloud mobile, social, big data, really transformative. And so I want to get your take. Docker, your company, is just exploding with buzz and growth in terms of developers, what's it like? I mean, a year ago, you're probably just the guy walking down the street, now it's like, hey, I want to meet you with Docker. So what's it like? Give us a share of some insight, personal insight. Oh man, well, this is actually my fifth startup, third as CEO, so I've experienced the opposite, so I'm not going to complain about what life is like right now, but it is, it's just insane in terms of the number of customers that want to work with us, large IT organizations that want to work with us, competitors bringing up, adoption, you name it, but it's an exciting time. We had Jerry Chen on theCUBE yesterday with Pete Stonestini, both two VMware alums and Steve Herod on yesterday, so it's good to see the VMware culture, but Jerry Chen has the hot hands, his first deal. Not bad. I want to make him very wealthy and very successful. Jerry, if you're watching, you'd be proud to hear that. He's behind you. He's a great guy, certainly a great investor, who's been really doing some due diligence on cloud and essentially, you know, Jerry kind of, and I'm going to paraphrase, this is my interpretation of our conversation over the years, is he's looking for the next Amazon in the enterprise. It's very clear Amazon's winning in the cloud game. All anyone who's doing any kind of startup as it goes to Amazon, it's a DevOps integrated stack, all that stuff, very developer friendly. Enterprise has requirements, there's certain, you know, compliance issues, there's certain things they need to support, they have legacy infrastructure, you guys come in with a container, not a new concept, but you had a twist to it. So I want you to explain to the folks first, what is unique about the Docker model that makes it so relevant today? And give them a quick tutorial on the key elements there. Sure. Well, I think just, you know, want to start by saying, you know, you think about where applications have come, you know, since the dark ages, like maybe eight years ago, right? I mean, it used to be that applications lived a long time, you know, people would talk about six months or even longer in terms of lifecycle, they were monolithic and they were deployed to a single server. And all three of those things basically out the window, right? So people want to develop rapidly and iteratively, they want to build applications from lots of loosely coupled components, and they want to run them on thousands of servers. And so somehow you need to get all of this complication of different languages, frameworks, components, versions running across, you know, from the developers laptop through testing staging to production and when you're in production, across clusters, moving to the public cloud, moving to the private cloud, moving to VMs, moving to bare metal, et cetera. And, you know, the intersection of all of that is a real mess. And so, you know, when we usually like to talk to newbies about this, we use the analogy of a shipping container, because there used to be a real matrix from hell in the world of shipping physical goods, where you'd have to worry whether you were shipping, you know, bananas next to somebody else's anvils, and things had to be unloaded and reloaded every time they went from a ship to a train to a truck to a crane. And now, you know, basically you put everything into the same container and it moves from the ship to the train to the truck to the crane and nobody cares what's on the inside if you're a ship captain or what's on the outside if you're a manufacturer. So, for us, we wanted to make the same thing happen in the world of applications where, frankly, takes longer to ship applications from one data center to another than take car parts halfway around the world. So it makes things a lot easier in terms of development. You put the container in there for the developer and a lot of the other stuff just gets taken care of. That's essentially what I think would, you know, use this analogy, right? What's on the inside of the container is invisible to everything on the outside and vice versa. So every dockerized application, you know, starts the same way, stops the same way, logs the same way, introspects the same way. So if you're a developer, you can be wild and crazy on the inside of your container. And you can stitch lots of containers together in a really easy way. And if you're an ops guy, they all look the same. So you can be automated and uptime and stable in all those goals that you talked about before. So this brought us to the EVO Rails thing we talked earlier about. Servers look more complicated now. It's supposed to be more simpler because with virtualization, all that's happened. That's what you're referring to. So all those guys innovating in the virtualization space, look at the container can do their business and get faster and more simpler and it doesn't affect the applications. Is that what you're saying? Yeah, yeah. Well, I'm saying that you can be, you can build the application environment that you want that's agile and rapid and reusing components and secure and everything else like that without having to know in advance what the infrastructure's going to be. If your idea of ideal infrastructure is VMware, great. If it's OpenStack, great. If it's, I don't want to deal with any of this, I just want to outsource it to Amazon, that's great too. And the key is of course that everything behaves consistently and also that your unit of measurement is something that's really lightweight as opposed to having to take an application measured in megabytes with a guest operating system measured in gigabytes and somehow like that heavy thing around. That's the DevOps ethos right there. App developers aren't experts at infrastructure. And they just want it to work. Yeah. Well, I mean, look, you hear DevOps and a lot of people have this wrong picture that DevOps means that you're going to get developers and ops guys to hug each other and say kumbaya and agree on a set of tools, right? And they're fundamentally different, developers, they like to break things and experiment and ops people lose their jobs when things break. And with Docker, you get this nice clean interface where developers can do whatever they want and at the end of the day, it doesn't impact ops and ops can be. All right, so let me break this down. Application developers can have fun, break stuff and go crazy and ops guys don't get fired. Yeah, you know that pretty much it. So, anywhere you're done, you've got it. Docker, home run, go public. There's the bumper sticker. So Ben, leaving it on the show, and we interviewed Pat Gelsinger in May and of course I asked him about Docker and saw articles, what a threat Docker was and containers were to VMware. Pat's answer at EMC World in May was, wow, who's got the best container technology? Well, we do, blah, blah, blah. And we said, John and I remember we were talking and we said we'd better have a better answer by VMworld. Well, his answer at VMworld was great, you know, you guys are, and it's fantastic for you guys. So first of all, congratulations. However, when you hear things like containers without compromise, hmm, compromise, whose containers are compromising? And we had Chris Wolf on today saying, you know, this notion of right once, run anywhere, it sounds great, it's conceptually right on, but in practicality, it's not really that simple, right? You got to, if you're going to run on a stack, you got to run on a VMware stack or an Amazon stack or a Google stack. So what is the compromise with containers? Is there one and what is that? You know, so I don't really think there's a compromise. I think that a lot of people have a mindset about building applications that was based on the notion that, you know, you have to have everything tightly coupled, right? And so you sort of build your applications in a way that makes assumptions about what's running underneath, right? And so if you do things in that way, yes, there are dependencies. If you take sort of the more modern 12-factor stateless approach towards applications, so today there are lots of applications you can write in Docker and you take that, you build that same Docker container in seconds and you can deploy it in milliseconds to any different infrastructure without modification or delay. Now as we sort of go from simple applications to applications composed from multiple different containers, if you go from stateless to stateful, as you go from the web companies who are all over us right now to the banks and the healthcare organizations that have security concerns, yeah, there's an evolution that needs to happen. But the good news is you take Docker containers and for certain environments, you deploy them inside of a VM, you get the best of both worlds. Right, so from a partnership standpoint, you're an arms dealer, right? I mean, so you've got your VMware, you know, deal option. Oh, we like to think of ourselves as neutral in Switzerland as opposed to arms dealers. Yeah, I know. Of course not, you can't say that. Neutral, arms dealers are neutral. Okay, but Switzerland, maybe not. Okay, maybe a bad choice of words. But so take us through the strategy, the partnership strategy, and where are you at? In terms of knocking down. Our partnership strategy starts with our user strategy and our user strategy says, users should be able to build the application management and development and build and ship environment of their choice without us telling them what the right infrastructure is. Right, so again, if your idea of ideal infrastructure is running in a public cloud, we've got relationships with Amazon, Mac space, IBM, you name it, right? If your idea of the ideal infrastructure is running on bare metal servers, running, you know, a full stack OS, we've got partnerships and we ship with RHEL and with Canonical and with a bunch of others. And if your idea of the ideal infrastructure is virtualized, we ship an open stack, we run in VMware, right? So for us, that's the right approach, right? Is to let the users decide what infrastructure they want and for us, just make it a really clean interface. And more to come, how about software, I guess? Oh yeah, we run in software. You do, okay. Yeah, so a lot of great announcements with IBM in general. And BlueMix and OpenShift and a lot of tabs as well, so. Great, yeah. Arms dealer. So who, so wait, I love it. No, it's great, great strategy, it's beautiful. We don't have like big Swiss bank accounts yet. So how did the VMware deal come about? Take us inside that. Yeah, well I think like most great things with Docker, it sort of starts at the technical level, right? We're, you know, radically open, you know, open source. And I think that they started seeing the same thing that we did, so you know, as we start going into banks and healthcare organization, government institutions, they say, hey, we love Docker. We want to start with Docker sort of in our software development life cycle, but we want to make sure that we can leverage 15 years worth of security and management tooling from VMware. And they started hearing the same thing, right? So it's sort of natural for us to talk and say, okay, listen, you know, they understand that Docker wants containers to run everywhere. And we understand that they want Docker to run well inside of VMware. So at that level, I think we're completely aligned. So you spent some time in the East Coast, I think, right? So you remember when Bill Parcells was, you know, running, whether it was the Giants or the Patriots, I remember he said, oh, hold on, let's not put him in the Hall of Fame yet, you know. So if people were saying that about Docker, well, those are the 42 employees that you've probably grown since then. Now we're 50 employees and a pet turtle, so, you know. You got the turtle, well, okay. Well, actually, he now weighs enough that when he walks across the keyboard, he depresses the keys. So I'm now officially the second worst builder, not the first. You've been feeding him. So where are you at with the company, its growth? I mean, why don't you talk about that a little bit? Sure. I mean, look, we're still a fairly small company. You know, fortunately, we're able to be much larger than we are in terms of the company itself. There are over 600 contributors to the Docker project. You know, and again, I pay salaries of 30 of them. If you go on to GitHub and you search for Docker-based projects, you'll find over 12,000 projects with Docker in the title. So, you know, that's everything from Chef, Puppet, Salt, Ansible to new things that are being built. If you go and you want to buy, or you want to get pre-baked components, pre-dockerized components, there are now 30,000 in our public index, right? And, you know, 250 meetup groups around the world so that as a company, we've managed, and as a project, we managed to get much bigger than, you know, the number of people would suggest. Okay, and can you talk about how you're going to make money? You know, interestingly enough, our investors ask the same question. So actually, to me, I think it's fairly straightforward. Look, there's a few models that work in open source. There's the Red Hat model where you charge for support and right now we offer commercial support. There's sort of the GitHub model where you have a hub, where you have a bunch of free services and people want to pay for things that are more enterprise-grade. And we have that, we have something called Docker Hub, which is a, you know, a center for collaboration and content workflow, focused more sort of on the development side of the house. And then we're introducing a set of functionality that will ultimately be more around the monitoring, management, and orchestration of containers that will be sold as a commercial product. All of the above strategy. Yeah, why not. Arms merchant and all of the above. You're a marketplace for developers. Yeah, yeah. Marketplace or... That's a big part of it, right? We have, right now we're bookstore. Right now we're library. We'll be a bookstore soon. So I got to ask you about the open source component. When we started digging in and looking at the flywheel of traction you had, certainly you've had traction from big one with developers, but now, what's happening is interesting. When I talked to some real high-end developers, I said, hey, you know, Docker's awesome, you know, obviously what's going on? Everyone is pointing to your community piece of it. Explain to the folks out there what's unique about it. It's not just open source. It's open source with a twist in the sense that there's almost an disincentive to share your Docker. So what does that mean? I mean, normal open source is, hey, I donate open source, I can use it and create some proprietary value and go over here and make money on it. What's different about the Docker open piece of it and why is it so successful? Yeah, well, so I think we're an open source project, but really I think of it as more of an open source ecosystem or an open ecosystem. And Docker itself is not just an application, it's a platform. So that not only do we have lots of people who are working on the Docker code base itself, but we have orders of magnitude greater than that in terms of people creating applications that run on Docker. And at this point, 12,000 different projects of tools that work with Docker that go all the way from the red hats and open stack and VMware down to little two person shops. And to me, that makes it really powerful because then you're not just a piece of code that can be copied and forked, right? It's, if you want to fork Docker, you're forking an ecosystem and you're leaving a lot behind. It's a lot, it's a folksonomy built into the open source ethos, which is sharing, so the more you share, the better it gets. That's right. And that means for us, we have to be really good stewards and be sort of radically open and welcome competition as well as collaboration. I think that takes the business model off the table because what happens is you get critical mass, the monetization downstream becomes less, more of a obvious thing versus trying to force it now. Is that the way you see it? If you try to force your business model now, you potentially foreclosing growth, right? Absolutely, look, I mean, you know, I could be a small proprietary company trying to knock on doors, trying very hard to have five customers. You could make some money doing that. Make some money. I would rather have millions of users and we have like 13 million downloads. And if I can get 10% of them to pay me, that's a much better model. Sort of dry cleaning tomorrow, make some cash and pay some money out. Well, right now we're, you know, our biggest money maker right now is Docker T-shirts, but you know, we're looking to change that. And I think that takes the pressure off a lot of these new models and David and I are talking about our engagement container, CrowdChats which we've been called, the Docker of social media, little plug there. I love that, yeah. But the business model is the issue. Twitter at the same problem, you know, and they said, hey, no, we want to win here with the product market fit first. Right. And you're funded so there's no real pressure. You got plenty of cash in the bank. You're not going to go out of business anytime soon, right? No, no, no, and listen, you know, we've got a platform, the platform is growing, it's organic, and we've already got a pipeline of some 200 Fortune 1000 companies that want us to... To take their money. Take their money. Take their money, yeah. And you know, on our side, we do want to take it. So... And do it in an elegant way. Provide some value. So we got that. So now let's get to some of the technical things. Windows support, Linux is there, but what about Windows? So, you know, we're starting with Linux and that's a pretty big area to cover. Next would be Unix just because there's some primitives that are really easy for us to use. And Windows is definitely on the horizon. There's actually some great work being done in the community. 32-bit is on the horizon, but not this year. Okay, got it. We've got some time. Now, the thing I brought up with Bill Fathers in the cube earlier, VMware and others like CTOs is the stateless applications certainly winning there with Docker. All the developers using REST and all the APIs are doing very, very well with Docker and very happy and it's flying off of the shelves. Stateful applications. Still, state is important. How do you guys look at that and what's on the roadmap? Can you share specifically how you're dealing with state? Sure, sure. I mean, so, you know, first I guess I'll tell you that there are plenty of people who are playing around with things like SAP and Oracle inside of Docker containers. So, there are a few different things that we do. We give you- Interesting. Yes, yes. I love it. You'd be surprised, right? But so, first of all, we're going more towards the model of multi-container apps, right? So you have your application in a container, you have your database in the container, you have your data in the container and so, in many ways, you can sort of put state in one place without having to worry about state in all the places. You know, Docker also changes the way people think about state because if the unit is quick to create and quick to destroy, a lot of your concerns about state also go away. And finally, you know, in the near term, you know, you want to free state and live migrate, use Docker plus VMware or Docker plus another VM. In the medium term, you'll see us adding things like live migration and high availability and all of those other good things that people ultimately need. Final word, just because we have to end this thing, we're getting the hook here. Share the folks out on your own words, this whole app tsunami that's really dictating the growth right now is also coming from the app side. What's your take and how would you summarize that marketplace? I mean, I think, you know, ultimately the developers are in control, right? And they're defining how things get built and increasing what they want to be able to do is they want to be collaborative and sort of as open-sourced as possible, right? They want to build multi-component apps where 90% of the components are pre-baked and they can focus on the things that are most critical and differentiating and they don't want to worry about infrastructure. And the preferred user experience from the developer standpoint is I just need infrastructure on demand, pretty much. I need infrastructure on demand and in case Docker we also say, you don't even have to worry about what the infrastructure is going to be. You know, you've got a unit that will run on any infrastructure. Ben Golff, CEO of Docker. We had Solomon on the queue but Red Hat Summit, congratulations, great company. I know you sold off .cloud and did some things there, getting focused on developers. Developer developers was Balmer's thing. You guys are apps, apps, apps. That's the model here. It's a DevOps, it's a DevOps cloud world. It's theCUBE. We'll be right back with our next guest after the short break. Docker inside theCUBE, breaking it down for us. Thanks so much, we'll be right back. Thanks a lot.