 Live from Vancouver, Canada, it's theCUBE at OpenStack Summit Vancouver 2015. Brought to you by headline sponsors EMC and jointly by Red Hat and Cisco with additional sponsorship by Brocade and HP. And now your hosts, John Furrier and Stu Miniman. Okay, hello everyone, welcome back to the day three of live coverage here at the OpenStack Summit. This is Silicon Angles theCUBE, this is our flagship program. We go out to the events and extract the civil noise. Talk to the expert, talk to the entrepreneurs, talk to VC, talk to whoever has data to share. We share that with you. I'm John Furrier, my co-host Stu Miniman. Our next guest is Brian Redbeard, Principal Architect Coro West, welcome to theCUBE. Hello, thank you, thank you, I'm glad to be here. Love the look, it's great. Redbeard, it's beautiful, very, very hipster, OpenStack looking, we love it. But a lot of action here and I want to get your thoughts before we get into some specific questions around the vibe and around the people here. I mean, you were just gave a talk yet last night, packed session, standing your moment, people literally on the floor. This is a telltale sign that there's real building going on. People are actually acquiring information. This is on the path to maturity, as we were saying in the intro. What's your take on that, what's your view of that? What's your big picture take on all this learning? It's active community, but a lot of change going on. Absolutely, you know, there's, when it comes to what we're doing, like we have a clear vision for where we want to take infrastructure and it just, it happens to be that there's a lot of outstanding questions at the community level and that's what folks are here to actually kind of solve and work through and it's nice to actually go where our users are and work collaboratively to figure stuff out together. So what's your take on the overall core OS mission, obviously, you gave a talk on OpenStack with Kubernetes, obviously Google is involved with you guys from a funding standpoint. What's the progress? What's going on? What's the state of the union? What's the landscape look like for you guys? Well, the big things for us right now are really targeting Kubernetes, kind of pushing adoption of AppC and then greater use of Rocket itself and really trying to continue with the overarching mission of changing how infrastructure is done and driving greater security on the server side of things on the internet. Yeah, so, so, so, Redbeard, one of the big discussions we're having at this show is, you know, what are the business models for open source? You're the first guest that we've actually had on from Core West, so I've had the chance to talk to you a few times, talk to Alex Pulvy, talk about kind of the Core West's mentality as to how you build, you know, projects and tools and, you know, how to put them out in the marketplace. So very much we are an open source company, we are an open company in general, like when right from the start we are present on IRC, we're very active on GitHub, you know, we try to be as present as we can to our actual users, you know, we do a lot of like community meetups and stuff, but what it means is that all of our development happens in the open, folks can really track what we're doing and we make ourselves available so that when folks want to kind of see where our vision is going, they can do a little reading of the tea leaves when they actually watch our Git repositories and get a good sense of what's coming next and that's why when it came to some of the stuff with like Kubernetes and Rocket, folks start looking at seeing, oh, I see that there's some commit changes happening on Rocket, this is where it's going to go and, you know, from a marketing perspective, it means that there's also just a little bit less that we need to do on that, because folks do have an overall sense of like where our infrastructure is going. All right, can you speak a little bit about the CoreOS involvement in OpenStack then, which projects are you guys heavily involved in and what are you seeing, because, I mean, Magnum is pretty new, Kala's there, you know, other pieces, there's the keynote, the discussion of Kubernetes and how that's expanding into OpenStack. Yeah, so just under the wire, I was able to get some commits in on the OpenStack marketplace, the Mirantis just rolled out, so, you know, we've been involved with that. For a long time, we were working with the folks over at Rackspace and HP on Ironic, Ironic utilizes the CoreOS image to get a lot of things done on that front, and, you know, we've been talking both internally and with the folks on the Magnum team about how CoreOS gets involved, like there's already been a few kind of proof of concepts where you can use both Project Atomic and CoreOS inside of Magnum today, so we're really trying to make sure that where it makes sense, we're present and active in these projects. You guys are optimizing on many levels, but can you share with where your focus on optimization is in terms of the efforts and the coding? Is it the secure cloud? Is it container delivery? If you guys had to optimize, and where are you guys focusing all your guns on? I mean, obviously there's a lot of stuff going on across the board with Linux, you know, the distribution piece of it, but what is that focus? What do you guys organize around to optimize first? Well, so, like a lot of open source projects, we have federated teams inside of CoreOS that work on specific things, so, you know, in the same way that you'll see a lot of interaction between like the Apache and the HA proxy teams at the open source level, internally, we have them separated, so like the rocket team is doing things and the tectonic team is doing things and then, you know, they do a sync up. So it's been continuing on the focus of security, like we've got Matthew Garrett from our team who's done a lot of work recently on Secure Boot and UEFI, and it's where we're trying to bring the overall security of the stack from being able to verify the bootloader to the image itself and eventually securing the containers as well and being able to validate that. Still at the community, what's the big reaction? I mean, obviously this is growing fast. There's a lot of traction across the board. Where are you guys seeing the collaboration coming in around CoreOS? What are some highlights? Can you share some data around where the collaboration is? Absolutely, so we just had CoreOS Fest, which was our open source kind of, you know, whereas some other companies have like their big commercial event, we wanted to have a big open source event. And it was a great opportunity because folks from data.com over at Salesforce came out and gave their story about how they're using the open source components. Intel and Supermicro came out. And that really also gives a view into where we're working with other companies and trying to contribute things upstream for folks to work on. Like the folks, the work that Intel did around optimizing SED to work on the silicon, much better was just out of this world. Like seeing the 10X performance increase and they're promising another 10X soon, so we'll see that. So, so Redbeard, you know, I think back to like how long it took Intel to optimize for virtualization. It took many years. How did this happen so fast with Intel? I was hoping Nick Weaver was supposed to come on. I think he's actually flying in today. And I know he was involved, but you know, how's CoreOS and Intel working so fast? Well, a lot of it just comes down to standing on the shoulder of the giants that is the Linux kernel team. Like the work that Linus Torvalds, Chris Wright, Greg Crow Hartman, and then all of the folks kind of also involved, they're really focused on working with hardware vendors to get hardware support in as quickly as possible. And it's just because we use an up-to-date kernel, we get all of these features really as soon as they're available to the kernel. You know, whereas other distros, they lag behind a little bit because part of their actual sales model is like having a more, I wouldn't necessarily say stable, but you know, having a kernel that lags behind so that they can control the feature set a lot more tightly, whereas we want to bring the features that are in the stable kernel to folks as fast as possible, because it is the stable kernel. And you know, having that strict contract around the API and the ABI, like that binary interface, means that your workloads are going to continue to work even when you use that newer software. So I got to ask you last year at Atlanta when we were doing theCUBE, I mean, all the rage was containers, Docker. And so Docker kind of slid in through the mind share and on the conversation cycles, you know, and so that buzz was rocking. I'll say Docker's got 95 million bucks and the container story's kind of been kicked around, but we're not hearing a lot about containers. It's not dominating the conversation. It seems to be a subtext, if you will. Kubernetes certainly has risen up as well. You mentioned that earlier. Why is that? Why is it because it's more established, more recognized containers and delivery with orchestration and with Kubernetes is now hot? Or is there a bigger messaging and or conversation going on on top of it? Can you talk about that as an architectural thing? Is it more of a systems architecture, services architecture? What is, why isn't it dominating? Does it already establish sub-element? Share your thoughts. I think that there's two key things to identify there. One is that if you look at the patterns of social media networks, there's a lot to be into it there. And also if you look at like the development of the iPhone, really it all comes down to content. Today, there's not a lot of open source upstream applications which are designed to run natively in containers. And that inherently has a slight limitation to organizations that are looking to do green field applications. They want to do that new development and so they know and they recognize that containers make sense for that. But until we see open source applications in the upstream that just, they treat containers as a first class citizen, it's going to be how do we then shoehorn this into containers. So that's why I make that comparison to social networks because it really comes down to having the network effect. Yeah, and some critical mass, some tipping point if you will. So you got to have some ubiquity. So people are recognizing, so I break that down, they're recognizing the importance of containers. But the native apps are not out there. So it's now being architected. So we are in that kind of pre-architectural figuring out for the apps. Yeah. So how many plays are there at Core West today? There's about 40. Yeah, so am I right? You guys have about 90 projects up on GitHub right now. You know, I had a great conversation last year on theCUBE with John Ankees from Rackspace. And he said, we hire great people and we let them work on whatever they want. And it was like, wow, how do we make that turn into a business? So the question I have for you, Core West, I mean, you know, SCD, Rocket, Fleet, Flannel, a whole lot more up on your Git page. How do you guys focus on that? How much of it are you guys driving? How do you bring the community in? And then we'll talk about Tectonic when we're done with that. Well, a lot of it comes down to our employees, we try to make sure that they're empowered to be able to work on the things that are interesting and really kind of feel out all the threads to see what makes the most sense. So that's why you do see so many repositories up there because we have an idea, we take it, we run with it, and then we see what makes sense in the long run. When it comes to that, you know, it just is a natural progression where, you know, I have an idea and I start looking at that. You know, the one that I kind of started a while ago was called Mayday because we wanted to see a good mechanism for, from a support perspective, users to be able to send telemetry data back, but that really just captures it in a way where they can unpack it and see where that goes. And you know, today that's modeled after some other patterns like sauce report and things, but it's just one example of where we take these little ideas and we, in the true UNIX model, try to have small tools that do that one thing well and then see where they go from a ubiquity perspective. All right, so then can we talk about the CoreOS of the business? Because in addition to making all these pieces, you guys are building a platform, tectonic, huge discussion at the CoreOS Fest. You know, how's that going? Where are you with customer deployments of stuff that they might actually pay for? Yeah, so right now we have a closed beta going on. It's very, very high touch. You know, we've specifically kept it to a small group of people, a small group of companies, so that I can interact with them on a weekly basis, and make sure that they're, that it's operating exactly the way that they expect it to. Because anytime you build software, there is that gap between what we thought would be useful and what the customer finds useful. So I'm there kind of as the reconciler to make sure that those actually converge on a central point. And it's really promising because we've got a really diverse group of customers in that beta program between financial, media companies, large data aggregation. So we tried to make sure that we had a good cross-section of folks with large workloads to be able to test out and ensure that this would be really good for those companies from a commercial perspective. What's your advice to folks out there that are getting in? Because a lot of in-migration coming into the community, which is great, you see some fresh blood come in. And it's not tire kickers either, these are real people, engineers. What's your advice then as they come in and gratiate into the community? Obviously the communities will open to your community and open stack in general. How would you advise them on the navigation, like where to get started, there's so much to jump into. So how does someone just get in, what's your advice? I mean you've been doing a lot of cool stuff so obviously that's going to attract some intoxication effect with developers. So what's your advice for the newbies out there coming in, serious players? You know, what I would say is in similar to where I said earlier that the Linux kernel is what is enabling this, there's been a very large pattern over the past 15 years with the kernel of folks coming into Linux but just treating it like it's Unix and it really does short shrift for Linux where if you're just coming in and expecting to be able to treat this like Linux, you're not really getting into some of the next generation features that are there. Like when we look at this, Google contributed a lot of the code in coordination with IBM back in 2005 that really laid the groundwork for containers. And 10 years on we're just getting around to doing that. So for the new folks and the folks who like have a strong architectural sense where they want to be able to work on that next generation, it's that you want to actually look at the kernel and understand what differentiates the Linux kernel from some of the more traditional Unix models because that is what's going to unlock these features for next generation. So really the creative pieces and the exercises to come in and understand where the leverage points are for the next generation apps. Exactly. And because it does take a little bit of time and the second thing I tease out of that is that, and this is always something we always talk about in theCUBE, is that it's a systems architecture kind of concept right now. There's a lot of really great work and the affinity from the developer engineer is a architectural issue, right? There's a lot of architecture going on right now. Yeah, it's something where the folks who are architects at large scale media companies or at large scale companies in general, they're going to be able to have the vision for where they want to get to. And by having an understanding of the individual pieces, they're really able to, like an engineer, kind of build that up one piece at a time and make sure that they're all going to interoperate in the way they need to. So one of the questions I have for you is if you look at kind of the whole cloud conversation in general, if I take something like OpenStack, where is the line where below this line, you might not want to do it yourself. You want to kind of go to the rack space, go to HP Helion, go to somebody else and let them do it. We've said on theCUBE many times, friends don't let friends build new data centers. So what can IT get rid of? But containers are one of those things that, whether you're really small or big, it might make sense for you. Developers are loving the containers in general. You guys work at like massive scales, one of the things you work on. So where's kind of your natural fit and where's that kind of, is there a low watermark or where's that look go? You know, it's one of those things where it's hard to hear, but it really does. The answer is it depends, but there is a bit more insight that we can get to on that. Because for each organization, you have the target of how technical you want to be with specific components. You know, if you are a hardware electronics manufacturer, you know, if you're the kind of giants of the industry, then you may have a team of kernel engineers that has an understanding. And at that point, just starting from like the core OS pieces and working up makes a lot of sense. Whereas if you're more of a small or mid-range business, you know, realistically, like you said, you don't want to be building out a data center. You don't want to be building out like your own set of APIs. And like it probably doesn't even make sense to be doing development on the OpenStack components themselves. So it does, like knowing where your core competencies are really kind of help get alignment on that. All right, so CoreOS is still relatively young as a company. What, as industry watchers and people watching you, what milestone should we look for? What are some of the kind of the goals and success factors that we can say kind of the end of 2015, you know, core OS is moving in the right direction, making money, you know, things like that. Well, so our major commercial flagship that we're launching, which is Tectonic, we're targeting for the second half of this year. And realistically, it's going to be hopefully early on that, but it means that when we hit the end of 2015, Tectonic will be the thing that folks will look at and go, okay, that has demonstrated that Kubernetes is a viable product for the enterprise and that core OS makes sense. That core OS really is solving real world problems that companies have today. Brian, thanks so much. Brian Redby, you're inside theCUBE. Thanks for sharing your insight and just your take on this year's summit of Vancouver. What's the vibe, folks that aren't here? Give a little taste and color of what's going on, just around the show. Share some color and insight. Well, when it comes to OpenStack Summit, it's one of those things where each time it gets a little bit bigger, each time it's moving from its infancy into adolescence, and it's going to be grown up here pretty soon. When you see companies coming in like Ericsson and Alcatel Lucent and all these real big players in the networking and large-scale computing space, it shows that it's proving itself as a technology that everybody wants to use. And more use cases on stage from customers doing live demos. So you're seeing kind of growing out of this POC. Exactly. Yeah, it is 100% real deal. And when you see all of the various companies that are investing in OpenStack, like a lot of open-source projects, you start it, you work through it, and it goes from the stage of getting things done to doing the things elegantly. And there are a lot of naysayers around OpenStack, but I think it's just natural to the fact that we're making that transition from components that just got the job done to now components that are maintainable in the long term. It's hard for folks to get their arms around the complexity of what OpenStack is, because it's such a broad, ranging, and diverse use cases. You got telcos to enterprises, the developers all kind of mashed in, so it's hard to put it into a little box and saying, you know, are these three layers when you have a service architecture model going on where people are actually creating their own blueprints, if you will. Exactly. That seems to be our takeaway from this week is that, you know, it's growing, but people really don't understand the animal that is OpenStack. What is this going to look like? The lion is the tiger, is the cat a dog? I mean, it's kind of crazy in a good way. So opportunities are there. Yes, it definitely is. So I got to ask you, the final question is, you mentioned reading the tea leaves, and so how would, how do people, you know, if you could share your secret sauce on reading the tea leaves, whether it's in Git or in the projects, how should someone coming in who's going to jump in for the first time and do or is immersing into the community, how do they read the tea leaves? What's your secret sauce? What's your, what's your technique? The easiest one, and this is going to be a little bit of a giveaway for the analysts and folks out there who are trying to get in on it, but it's less the Git repositories themselves and more watching the issues in the repositories, because when you are seeing the issues that folks are raising or the feature requests that are being done, and then actually watching the comments on those and seeing what gets closed out, it really gives you a clear vision for where a project's going and what folks are asking for from the project. It's great metadata. Yeah, actually engagement there, you know, right, if you're not finding problems, there's probably not really getting used to any real environment. Exactly. So that's a great one. Brian, Redbeard inside theCUBE sharing his wisdom, talking about CoreOS and all the traction, all the opportunities out there. Great time to jump into OpenStack if you're an architect, you're an engineer. A lot of build out going on, it's exciting times. This is theCUBE, of course, sharing that data with you. We'll be right back after this short break with more guests here. Day three coverage here at OpenStack Summit in Vancouver, British Columbia. We'll be right back.