 Live from Los Angeles, it's theCUBE, covering Open Source Summit North America 2017, brought to you by the Linux Foundation and Red Hat. Okay, welcome back, and we are live in Los Angeles for theCUBE's exclusive coverage of the Open Source Summit North America. I'm John Furrier, my cohost, Stu Miniman. Our next guest is Steve Posty, who's the director of developer advocacy for Red Hat, CUBE alumni. We last spoke at the Cisco DevNet Create, which is their new kind of cloud native kind of approach. Welcome back. Thank you, thank you, glad to be here. We're here at the Open Source Summit, which is a recognition of all these kind of Linux con, all these things coming in, it's a big 10 event, love the direction, validation to what's already happened, and the recognition of open source, where Linux is at the heart of all that. Red Hat, I'll see you guys, are the Linux standard and gold standard, but it's more than Linux now. We'd like to give it that way. There's more than Linux on top of it now, so it's the recognition that Open Source is so much more. For sure. I mean, you could even see, who would have thought that there would be a whole huge hubbub about Facebook doing a separate license for the React libraries and all the interactions with Apache, the Apache Foundation, like the Open Source is so much, it's the mainstream now, right? Like basically, it's very hard to release a proprietary product right now and come up with some justification about why you have to do it. And why, and can it even be as good? Right. So there's two issues, justification, and then performance. Yeah, quality, all that stuff. And also, customers' acceptability of that. Like, oh wait, you mean I can't actually see the code, I can't modify the code, I can't pay you to modify the code and share it with everybody else? I think customers have come to a whole, the users of Open Source stuff, it's so permeated now that I think it's hard to be in the market without, I mean, look at everybody who's here, right? Some of the people that are here were some of the biggest closed source people before. Microsoft is here. Exactly. IBM is here, oh, IBM's always been open, they were big on Linux early on. But now you're seeing the ecosystem grow, so we see some scale coming, but there's still a lot of work that needs to get done, we see greatness like Kubernetes and serverless, offering great promise and hope for either multi-cloud workflow, workload management, all this cool stuff, but there's still some work to be done. For sure. What's your take on progress? Where are we? What's some of those under the hood things that need to get worked on? So progress, I think the funny part is our expectations have changed so much over time. So Kubernetes is about a little over two years old, and we're all like, oh, it's moving, so why is it not doing this, this, and this? Whereas like if this was like 10 years ago, the rate at which Kubernetes is moving is phenomenal, right? So basically every quarter there is a new release of Kubernetes, and we basically, we build OpenShift as a distribution on top of Kubernetes, and so we're delivering to our customers every quarter as well, and a bunch of them are like, this is too fast. This is too fast. Like we can't integrate all these changes, but at the same time they say, but don't slow down. Because oh, next release, we're going to get this thing that we want, and we know we want to go to that release. So I think Kubernetes definitely has more growing room, but the thing is, how much it's already being seen as like vstand, it's the way I like to talk about it, and I'll talk about this in my talk later, is I think for Red Hat, Kubernetes is the cloud Linux kernel, right? It's the exact same story all over again. It's this infrastructure that everybody's going to build on. Like now there are people who are standing up OpenStack on Kubernetes, or on OpenShift, right? So basically saying, I don't want to install and manage my OpenStack, it's too difficult. Give me some JSON and some components, and I'll just use Kubernetes as my operating plane. Yeah, we saw Kubernetes right out of the gates, Stu and I were at the first KubeCon. We were the president of creation, and just on the doorstep of that thing just unfolded, and we saw the orchestration piece as huge, but I want to get your take, if you can share with the audience why Kubernetes has taken the world by storm, why is it so relevant? What's all the hubbub about with Kubernetes? Share your opinion. Okay, so remember, okay, so this is a red shirt, and remember I work at Red Hat, so this is obviously a biased opinion. I don't want, I want to be upfront about that. I think there were a biased opinion. Right, well as opposed to a neutral opinion, right? Like we definitely, so, I'd say that in front of my audience is just so that I can go do your own research, but from my perspective and what I've seen in the marketplace, there was a lot of orchestration and scheduling out there, then it kind of narrowed down. There's three players, I would say, right now. The three players, I'll end with Kubernetes, but I just started with it. There's Kubernetes, there is Mesosphere, and there's Docker Swarm. I see those as being the three that are out there right now. And I think the reason, so I won't talk about the others, but I see those, why Kubernetes has one, is one community. So they have done a great job of being upstream, working with all people, being a very open community, open to working with others, not trying to make things just so it benefits Google's business, but to benefit everybody. The other reason is the size of that community, right? Everybody working together. The third is, I think they, so some of it's luck, right? I mean, everything. Timing is everything. You're on a wave and you're on your board. No, big wave comes, you surf it, right? I mean, that's what. That's exactly, so I think what happened with Mesosphere is they're a great scheduler. And a lot of people said they are the best scheduler to start with, but they didn't really focus on containers to start with, and it seems like they missed, like Kubernetes said, no, it's all about containers, and we're going to focus on container workload, and that's right where everybody else was. And so it was like, I don't want to write all that extra stuff in Mesosphere. I want to do it with Kubernetes because that's containers. And so that's the bit of luck lining up with the market. So I would say it's the community, but also recognizing that it's about containers to start with and containers are kind of taking over. Take us inside containers. You're wearing a shirt that says Linux's containers on the front. Our audience can see the back, it says containers are Linux. Exactly. Of course, Red Hat heavily involved. You're in the weeds like dealing with things that we're doing to make stability of containers, make sure it works in other environments. Tell some of the things you're working on some of the projects in the light. So some of the projects I'll be showing today is, one is based off of OpenSCAP, OpenSCAP. It's another open consortium for scanning for vulnerabilities. We've written something called Atomic Scan. So it can take any OpenSCAP provider, plug it into Atomic Scan, and you can scan a container image without having to actually run it. So you don't actually have to start it up, it actually just goes in. The other thing I'm going to be talking about today is Builda, which this is part of the CNCF stuff. This is the ability to actually build a RunC-compatible container without ever using Docker or Mobi. The way, and it's a totally different approach to it, what you basically do is you say, I want this container from this other container, or from Flank, then you have a container there, and then you actually mount the file system. And so rather than actually booting a container and doing all sorts of steps in the container itself, you actually mount the file system, do normal operations in your machine like it was your normal file system, and then actually commit at the end. So it's another way for some of our customers that really like that idea of how they want to build and manage containers. But also, there's a bunch more, like there's Cryo, which is the common runtime interface and the implementation of it so that Kubernetes now can run on an alternative container technology. This is Cryo, it's agnostic. If you look at Kelsey Hightower's latest Kubernetes in anger, I think, or Kubernetes in the pain of the hard way or something, his latest is building it all on Cryo. So rather than running on Docker, it runs all your container running, it happens on Cryo. So not that I'm not trying to say, well, of course, I think it's better, but I think the point that we're really seeing is by everything moving into CNCF and the Linux Foundation and getting around standards, it's really enabled the ecosystem to take off. Like Tectonic and CoreOS have done that with Rocket. We're going to see a lot more blossoming. The fertilizer has been applied from back from our... Yeah, and the CNCF, I'm two years old, it's just, I mean, there we have the fertilizer down, big time to get the manure, then all the thousand flowers are blooming from that. Yeah, like between Prometheus, I mean just Prometheus, Istio, there's just, I can't even see track of them. So Steve, you were talking earlier, customers are having a hard time with that quarterly release. Yes. How do they keep up with all these projects? I mean, we rattled through all of them. You've got them all down pat, but the typical customer, do they need to worry about? What do they have to focus on? How do they keep up with the pace change? How do they absorb all of this? Okay, so it's highly depends on the customer, right? There are some customers, or not our customer, just I'll just say users, who are advanced enough on their own, who they're out there basically just, they're consuming the tip of what's coming out of CNCF, right? All that stuff and they're picking and choosing and they're doing that all. For Red, had a lot of our customers, I like all that technology. You're our trusted advisor. When you release it as a product and I know I can sit on it for three years, because you'll support it for at least three years, maybe five years, then I'm going to start to consume it and you'll actually probably put it into a more usable form for me. Because a lot of the upstream stuff isn't necessarily made direct for consumption. How are you guys dealing with the growth prospects? We've been talking about this all morning. It's been a big theme of this show, is not only just the renaming of the variety of different advanced Linux Con, but Open Source Summit is an encapsulation of all the projects that are blossoming across the board. So, the scale issues. And as a participant, Red Hat, your biased opinion, but you're also incentive and you guys are active in the community, the growth that's coming is going to put pressure on the system. It may change the relationship between communities and vendors and how they're all working together. So, again, to use the river analogy, there's a lot of water going to be pumping through the system, and so how's that going to impact the ecosystem? Is it going to be the great growth that could flood everything? Is there a potential for that? I mean, you're an ecosystem guy, so the theory is there. It's like Jim stepping up with the Linux Foundation, talked to him yesterday, he recognizes it, but he also doesn't want to get in the way either. So there's a balance of leadership that's needed. Your thoughts? I mean, I think one of the things, so I mean, Linux kernel has its benevolent dictator, and that works well for that one community, but then you'll see something like Kubernetes, where it's much more of a community base. There is no benevolent dictator for life on Kubernetes, right? I think one of the nice things that the Linux Foundation has done, and which Red Hat has acknowledged is, let the community govern the way that works for that community. Don't try to force necessarily one model on it. In terms of the flood part that you were talking about, I think if you want to go back to rivers, there's cycles in terms of 10-year floods, 100-year floods. I think what we're seeing right now is a big flood, and then what'll happen out of this is some things will shake out and other things won't. I don't expect every vendor that's here to be here next year. And find the high ground. I mean, the numbers that Jim shared on this opening keynote is, by 2026, 400 million libraries are going to be out there, versus today, 64, roughly million. Right. You know, the ad from Cisco, things that's understated, but now there's more code coming in, more people. And so a new generation is coming on board. This is going to be the great flood and open source. I also think it's a great opportunity for some companies. I mean, I'm not high enough in Red Hat to know what we're doing in that space, but it's also a great opportunity for some companies to help with that. Like, I think that's one of the other things the Linux Foundation did, was set up the JavaScript Foundation, right? That is a community that... But that doesn't have no JS. It's a little bit separate. No, I know, but I'm talking about, but if you think about the client-side JavaScript, forget no, just think about client-side JavaScript and how many frameworks are coming up all the time in new libraries. That's a challenge. So I think that actually, that community could be one that could be good to maybe gain some lessons from as things happen more in open source. I think there are other open source communities. I'm wondering, like, no... But the feedback on the JS community is that there's a lot of challenges in the volume of things happening. And that's coming for us, though, right? That's what's coming, that's what's going to come for this larger ecosystem. So I think maybe there's market opportunity, maybe there's new governance models, maybe there's, I mean, this is where innovation comes from, right? There's a new problem that's come, it's a good problem. Your next point of failure is your opportunity to innovate. Exactly. And it's a good problem to have, right? As opposed to we have too few projects and no one really likes them. Instead, now it's like, we've got so many projects and people are contributing and everybody's excited. How do we manage that excitement? So another dynamic that we're observing, and again, we're just speculating, we're pontificating, we're opining, is fashion. Fashionable projects, never fight fashion. My philosophy is in marketing, don't fight the fashion. CNCF is fashionable right now. People love it, it's popular, it's trendy, it's the hip, new nightclub, if you will, in open source. Other projects just as relevant. So relevance and trending sometimes can be misconstrued. How do you guys think about that? Because this is a dynamic, everyone wants to go to the best party. There's a few missing out. I'm going to go check out Kubernetes, but also relevance matters. Yes. So I've seen this discussion internally and engineering all the time, when we're talking about, because OpenShift is trying to build a real distribution, not like, oh, here's Kubernetes, but a real distribution, like when Red Hat ships you the Linux, gives you Linux, we don't just say, here's the Linux kernel, have a good time. We put a whole bunch of stuff around it and we're trying to do that with Kubernetes as well. So we're constantly evaluating all the, should we switch to Prometheus now? When's the time to switch to Prometheus? Oh, it's trending really hot. Oh, but it doesn't give us the features. It's a balance. It's going to have to come down again. I hate this. It's a community. People vote with their code. So if something has traction, you got to take a look at it. That's basically what you're saying. But I would say, and this has been going on for a while and I've seen other people talk about it, if you are the lead on an open source project and you want a lot of community, you have to get into marketing. You can't just do- You got to market the project. You got to, and not in the nasty term of market, which is like, I'm going to lie to you and like a lot of developers think about like, oh, I'm just going to give you bullshit and lie to you and it's not going to be helpful. No, market in terms of just getting your word out there so at least people know about it. Lead with all your shopping. Socialize it. Yeah. I mean, that's what you got to get it. So there is a lot of chatter now. It's like, how do you get it noticed as a Twitter person, right? You have to do something. It's the same. It's going to be more like that for open source products. We're doing our share to kind of extract the signal from some of the noise out there. And it's great to talk to you about it because you help get the perspective and certainly red hat, you're biased. That's okay. That's your bias. Now take your red hat off. Okay. Hat off. Take your red hat off. Right. Put your neutral hat on. Okay. There's an observation of open source summit. I'll see the name change kind of significance in the sense that's a big 10 event. This event here, what's your thoughts on what it means? Come on, Steve. You've got a PhD in ecology so we want some detailed analysis as to how this all goes together. I mean, it's good marketing, open source summit, good name change, a little bit broader. I'm actually glad for it. So I've gone to some of their smaller events and I actually like this because it was hard for me to get to the smaller events to get quite enough people. If you, like this actually builds a critical mass, right? And more cross fertilization, right? So like it's much easier for me to talk to containers to car people, like, because the automotive Linux is here as well, right? Can't avoid people, you see them in the hallways. You can say, hey, let's chat. Let's talk about that stuff. Whereas in this small, so, you know, this is more about conferences. There's a good side and a bad side to everything. Just like I tell my kids, when you pick up a stick, you have to also pick up the other end of the stick. You can't just have like, oh, this is the great part, but you don't get the bad part. So the great part about this, really easy to see a lot of people, see a lot of interesting things that are happening. Bad part about this, it's going to be hard. Like if it was, if this was just CNCF, everybody here would be CNCF. All the talks would be CNCF. It's like a, you could deep dive and really go. So I think this is great that they have this. I don't think this gets, should get rid of. Smaller, more focused events. Well, the KubeCon, or KubeCon, the CNCF event in Austin will be there. That will be CNCF. Exactly. So that's, I'm glad they're still doing that. So they're going to have the satellite events. And I like this the best way to do it. I think, I think a big 10 event like this is good because this is small even today, but with the growth coming, it'll be convention hall size in a matter of years. Right, exactly. And the fact that you've made it into a big, and the fact that you've made it into this cohesive event rather than going to somebody and saying, hey, sponsor these five events. It's like, no, sponsor this one big event and we'll get most of the people here for you. It's also a celebration too. A lot of these big 10 events are, because education, you can get online, all kinds of collaboration tools online. But when you have these big 10 events, one of the rare things is it's the face to face people centric in the moment engages, engagements. So you're learning in a different way. It's a celebration. So I think open source is just too important right now that this event will grow in my opinion. For sure. Thousands and thousands of people. I mean, there are a thousand at some point easily. Yeah. I mean, I think definitely it's theirs to lose. Let's put it that way. I think the way that... I'll tell that to Jim's family. Don't screw it up. Don't screw it up. I think the way that you could almost think of this is a OSS con, right? Is it a comic con? This is like, this can become OSS con, right? Which is like, they should probably... In the same way that the Kubernetes Foundation works and grows with a lot of other people, it'd be great if they could bring in other foundations as well to this. I know this is being run by the Linux, but it'd be great if we could get some Apache in here or some Eclipse in here, right? I mean, that would just be... Total home run. Those foundations, bringing it in. I would truly make it an open-source summit. Yeah, exactly. As opposed to the World Series that's only in the United States. Yeah. Although, you know, I was at a hotel recently and they had baseball on. It was Little League Baseball, though. Their World Series is actually... Little League World Series is actually the World Series. It's a global World Series. Yeah, like, they're... It's the world. Yeah, as opposed to, you know, the MLB, right? That's it. Anyway. Steve, great to have you on. Any final thoughts on interactions you've had, things you've learned from this event that you'd like to share and pass on? No, I just think this space is great. I'm really excited to be in it. You know, I'm starting to move a little bit more up to the application tier at my role at the company and I'm excited about that to actually... So I've been working down to the container tier and orchestration tier for a while and now I'm excited to get back to like, well, now let's actually build some cool stuff and see what this enables on the up... And DevOps is going mainstream, which is a great trend. You're starting to see that momentum beachhead in the enterprises, so... One takeaway message for microservices people. Please put an ops person on your microservice team. Usually they start with like the DBA and then they say the middle person and the front end people. I really want to make sure that they start including ops in your microservice team. And why is that? What'd you learn there? What's the reasoning? Well, because if you were going to do microservices, you're going to be... The team's going to end up doing ops-y work, right? And it's kind of foolish not to bring in somebody The reason you want all the team together is because they're going to own that and you also want them to share knowledge. So if I was on a microservice team, I would definitely want an ops person teaching me how to do ops for our stuff. I don't want to reinvent that myself. You got the right core competencies on that team. It's like having the right people in the right position. Exactly. A skilled player. Okay, we're here live in Los Angeles to see the cubes coverage of Open Source Summit in North America, I'm John Furrier. Stu Miniman, more live coverage after the short break.