 Live from Austin, Texas, it's theCUBE. Covering DockerCon 2017, brought to you by Docker and support from its ecosystem partners. Hi, I'm Stu Miniman and joined with my co-host for the second day of theCUBE's program, Jim Kobielus, really excited to have not only the founder of Docker, Solomon Hikes, he's also the CTO, Chief Product Officer, did some keynotes here, all over the place. So Solomon, thank you so much. Thanks for having us. Congratulations on all the progress and welcome back to theCUBE. Thanks a lot. It's a lot of fun. All right, so, you know, so many things to talk about, but let's start with you. How you doing? You know, I'm sure, you know, there's so much that went into this week. You know, what are you most proud of? What are you most excited about these days? Boy, yeah, where to start? I mean, the cool thing for me about DockerCon is I focus on the keynote, right? We just package up a nice story, try to explain what we're doing, what we're going, and that's a pretty massive team effort. I think it's 30 of us for months preparing, you know, deciding what we want to talk about, working on demos, pulling all-nighters. So it's just really fun to see a keynote go from nothing to a really nice, fun story. And then I get to show up and discover all the other cool stuff. So I'm like everyone else. I just marvel at the, you know, the organization, the crowd, the energy. I just, I'm a happy camper right now. So it's interesting some of that dynamics in the industry is to, you know, okay, what's the important part? Who contributes to what? What fits where? Two years ago, we kind of had, you know, the hugging out as to, you know, the runtime and, you know, have the open source foundation step in. You know, big thing at the keynote yesterday. You know, two big things. It was a Moby project and Linux kit. Can you maybe unpack for our audience a little bit? You know, what is Docker, the company? What's the open source? Who are, you know, some of the main players? And, you know, it was the whole keynote so we don't have time to get into it. But, you know, what's real and what's there? Yeah, so you're right. That was the big announcement, the Moby project. So, you know, the, basically in an Excel, we launched Docker and we made it a product and an open source project all rolled into one. We just kind of adopted this hybrid model, building a product that would just help people be more efficient, developers and ops. And at the same time, we would develop that in the open. And that really helped us. And it participated in the appearance of this just huge ecosystem. And it was a big decision for us. And over time, both grew. Docker grew as a product and it grew as an open source project. And so over time, we had to adapt to that growth. And on the open source side, that meant gradually spinning out smaller projects out of the main one. So now we have dozens of projects. Literally, we've got Container D, we've got SwarmKit, we've got InfraKit, we've got all these components. And each of those is a project and we integrate them. And what we're doing now is we're completing that transformation and making sure there's a place for open source collaboration. Free for all, openness, modularity, try new things, move fast, break things maybe. And then there's the products that integrates, takes the best parts, integrates them together, make sure they're tested, they're solid and then ships that to developers and customers. So basically, we're saying Moby is for open source collaboration, it's our project. And all of it. And Docker is the product that integrates that open project into something that people can consume that's simple. So it's two mutually, it's two complementary parts to our platform. And could you talk a little bit about that there's kind of that composable nature of what you're building there. There's what Docker will build from it. And I think you've got a couple of examples of some of your partners, what's going to happen in the cloud, what's going to happen to some of these other, walk us through one of those. Yeah, so everything about Docker is modular. So really if you install Docker for your favorite platform, whether it's the Mac, Windows, your favorite cloud provider, et cetera, a Linux server, you're actually installing a product that's an assembly of lots of components. And like I said, these components are developed in the open and then they're assembled. And now with the Moby project there's a place to assemble in the open, start the assembly in the open so that other companies, a broader ecosystem can collaborate in the assembly and kind of experiment with how things fit together. The really cool thing about that is it makes it way easier to port the platform to expand it and to customize it. So if you're a cloud provider and you see all the pieces and you think, well, I could optimize that, I could add a little bit of magic to make it work even better in my cloud or in my hardware, then you can do that in the open. You can do that with the community. And then you can partner with Docker to test it and certify it and distribute it as an easy to use product. So everything can go faster. Yeah, so you mentioned open a lot there. Does that mean that Docker is now closed? You know, there's certain people that are very dogmatic when it comes to open for us. So maybe you can parse that for us. Yeah, so I think it's the same people that were complaining before that we were confusing our products and open projects. And so we think of ourselves as having a lot to learn and there's an ecosystem that's made of a lot of people and companies and projects that have had a lot of experience with openness in the past. So we spend most of our time listening, figuring out what the next step should be and then taking that next step. So people told us clarify the relative place of open source collaboration and your product. And that's what we did. Now I'm sure someone's going to say, well, I preferred it before. Well, you know, we just have to at some point choose. So the key thing to remember is Docker does everything in the open and then integrates it into a product that you can use. And if you don't like the product, if you want an alternative, then you still have all the pieces in the open right now. So I would say no, not only is Docker not going closed, we're actually accelerating the rate at which we're opening up stuff. Yeah, I mean, personally, I felt it was a nice maturation of what you've done before, which was batteries are included but swappable. But we've kind of taken the next step. It reminds me of those cool little science kits my kids get where it's like, oh, okay, I could pre-build it or I can do it or I could do some other things. That's a good, we've used that tagline. It used to be Docker has batteries included a bit swappable and you can make other batteries and we'll swap them into the product. We'll decide what's in there. Now everyone can do the swapping. It's a big free for all. And honestly, it's fun to watch, right? Is there any piece of Docker in the project outside of core Docker that Docker the company will refrain from building, will rely on ISVs to build or will Docker the company get involved or reserve for itself the latitude to get involved in development of more peripheral pieces of the overall project going forward? Yeah, so we spent a lot of time thinking about that. And honestly, there's so many different constraints. We've just decided we're going to follow the users, follow the customers. So we just want a platform that works and solves people's problems. And that's the starting point. And from there, we work out the implementation details, what technology to use, the order in which to build things and also what makes more sense in the core platform and what makes more sense as an add-on. So it's kind of on a case-by-case basis. Is there a grand vision document or functional service layered architecture that all of these components of the project are implementing or enabling? I mean, will Docker as a project ever be complete or will it always be open-ended? Will it constantly evolve and broaden or possibly broaden and scope continuously and definitely? Well, I think if you look at the Moby project on the one side with experimentations and all the building blocks, I think that's going to just continuously expand. And really, openness is all about scale. There's only so much one company can build on their own. But if you show the ecosystem you're serious about really welcoming everybody and allowing for different opinions and approaches, then honestly, I think there's no limit to how large that project can scale. I think Moby can go into tens of thousands of contributors as open source becomes easier and more accessible, which we're really working on. I think it could go into hundreds of thousands. That's going to take a while. So that will, I think, never end growing. I think Docker, the product, the company, the reason we've been so successful is that we've worked really hard to focus and be disciplined in what problems we want to solve. So it's a more iterative approach. We would rather solve less problems that solve them really, really well so that if you're using Docker for developing or running a production, you're really delighted. They're just every detail kind of fits together. So there's a roadmap, of course. We're going to do more and more, but we don't want to rush trying to do everything. Solomon, great progress on all of these pieces. So I've got the tough one for you. In the last year or so, Kubernetes has really exploded out there. Lots of your ecosystem is heavily using it. Is it that Docker Swarm and Kubernetes will just be options out there? I look at Microsoft Azure, and they're very supportive of both initiatives. Many of your partners are there. How should, how do you guys look at that dynamic and how would you like people to think of that going forward? Yeah, so I think it's actually, it's a great case study of why we're transitioning to this open project model with Moby. The whole point is that at any given time, Docker, the product will not be using all of the building blocks out there. It's just not possible. There's too many permutations. So we have to choose, right? And one of these building blocks is orchestration. And a year ago, when we decided to build an orchestration, we had really specific opinions on what it should look like as product builders. And we looked around and we decided, okay, it needs to be a new kind of a building block. So we built Swarm kits for our own use. And we integrated it. And now that there's an open project for collaboration, we're throwing Swarm kit in there so that everyone can modify it, extend it, and also replace it with something else. So I think the big change now is if you look at something like Kubernetes or Rocket as a container on time, or honestly I could make a super long list of all the components out there that are really cool and we don't use in Docker, now you can combine them all in Moby in custom assemblies. And we actually demoed that on stage yesterday. We showed taking some pieces from Docker and some pieces and taking Kubernetes as a piece and like plugging it together and saying, look, there you go, weekend project. I think we're going to see a lot of convergence and reuse of ideas and codes, especially in the orchestration piece. I think over time, the differences between Kubernetes, Swarm kit and others will really diminish. And we'll just integrate the bits and pieces that make the most sense. So I don't really think of Kubernetes as a competitor or a problem. I think of it as another cool component in the Moby ecosystem. And yeah, I think it's a lot of cool stuff. Absolutely, and I tell you, the Kubernetes community, it's just so thrilled that Container D is now open source. It really kind of solves that issue and really it hasn't been something I've heard a lot coming into the show. It was one of the themes we wanted to look at and it hasn't been something that is like, oh boy, fight war, anything like that, which is, hey, congrats on that. So I want to come back to turn back to your roots there. When I think about DotCloud to Docker, it was a lot about the application modernization. Fast forward today, and I mean, Ben's up on stage talking of the journey. How do we take your legacy applications and wrap them in, what do you think about that kind of progression? We kind of like that spectrum out there to help customers at least partially and be able to make changes. But I can't imagine that's when you started Docker, that that was one of the use cases that you really thought you'd use. So what surprised you? What's changed now? You build things, what do you see from customers? So actually, you'll find this surprising, but this actually was a use case that we had in mind from the very beginning. And I think that was lost in the noise for the first few years in the life of Docker because it became this exciting new thing. Come on, cloud native, cloud native. Yeah, exactly. You know, Docker is a huge developer community now and we spend a lot of time making it great for devs. The truth is, I used to be a sysadmin, I used to be on call, I'm an ops guy first and we learned how to help developers. Developers are the customer. And the Docker came out of our ops roots and then it evolved to help the developers. And that's something that's now lost in the noise of history. And it's a really pragmatic tool and it's built to solve real problems. And one design opinion we baked in from the beginning is that it has to allow you to do things incrementally. If Docker forces you to throw away what you have just to get the benefits, then we screwed up. The whole point is that Docker can adapt to what you're doing. So for example, you'll see a lot of details in how Docker's designed to allow for stateful applications to run in there, to allow for your own network model to fit before Docker, all the container resolutions, all the paths required you to change your app. Even things like port discovery, you have to adopt, you have to change the source code. And Docker did not require that. It gives you extra things you can do if you want to go further. But the starting point is incremental and honestly, I'm really glad that now that's resonating. We're reaching that point in the community where there's a lot of people using Docker interested in that. Because for a few years I was worried that that would be missed in the noise of early adopters that don't mind rewriting everything. So really from the beginning, Docker was not just for cloud native, microservices, 12 factor, et cetera. So I'm personally as a designer of products, as a pragmatist, I'm just happy that we're there. How do you see Docker evolving to support more complex orchestrations for data? For hybrid data cloud environments, private and public, you got the likes of Microsoft and Oracle and IBM as partners and so forth. And they have these complex scenarios now they're customers, a petabyte scale and so forth. What do you see that going? The data, the persistence, the storage side of the containerization under Docker going. I think there's a lot of work to do. I think over time we're going to see specialized solutions for different uses of data. Data has such a big word. It's like computing. So just like computing now is no longer considered one category, but it's specialized. I think data will be the same. And I think it's a great fit for this modular Lego approach to the Docker ecosystem. So we're going to see different approaches to different data models. And I think we're going to see a lot of modularization and a lot of different assemblies. And again, I think a lot of that will happen in Moby and we'll see a lot of cool open stuff. We ourselves are facing a lot of data related questions and requests for customers. There's stuff in there already. You've got data volumes, right? And I think you're going to see a lot more on the data topic in the next year. Yeah, like containerization of artificial intelligence and deep learning and all that. Clearly that's terra incognita so far because, yeah. We're seeing a lot of really cool machine learning use cases using Docker already. OpenAI is all on Docker. We watch what they're doing with great interests. Are you a member of that consortium? Let's say friends and family. Okay. Yeah, so OpenAI came out of the white Combinator ecosystem and Docker is a white Combinator company. So we spend a lot of time with them. I think AI on Docker is a really cool use case. I'm a big fan of that. Cool, all right, let's do it. Yeah, so Solomon, you know, unfortunately we're running low on time. Last question I have for you is, there's so many things we can do with Docker now. And I saw like, here's a bunch of the use cases. It's like, oh, I can run lots of applications. I mean, everything from Oracle's in the store now, things like that. You know, what is the, you know, kind of quick win when you're talking to customers and that gets started? What's the thing that gets them the most excited that impacts their business the fastest? You know, it's... And it never comes down to one thing. Yeah, honestly, we keep talking about Lego. I think it's like asking what's your favorite Lego toy. I think we're maturing in the model. I think Lego is just the perfect analogy because it's a lot of building blocks and there's more and more, but there's also the sets. And so we're, I think we're consolidating around a few different sets. And you know, there's maybe a dozen main use cases and we're seeing people identify with one and then we're helping them see a starting point there. Like here's a starter set for your problem and then it clicks. All right. So yeah, I hear that and I can't help but think back. You're the big green platform that all my Legos build on. So I can have my space stuff. I can have my farm set. You know, maybe the Duplos don't quite fit on it, but you know, yeah, it's the platform, you know, helping me to modernize a lot of what we're doing. All right, Solomon Hikes, always a pleasure to catch up. Congratulations on all the progress here and look forward to catching up with you the next time we connect. We'll be back. Jim and I will be back with lots more coverage here from DockerCon 2017. You're watching theCUBE.