 Live from Austin, Texas. It's theCUBE, covering KubeCon and CloudNativeCon 2017. Brought to you by Red Hat, the Linux Foundation, and the Kube's ecosystem partners. Welcome back everyone, live here. This is the Kube's exclusive coverage live in Austin, Texas for CloudNativeCon and KubeCon with the Linux Foundation. I'm John Furrier, the founder, SiliconANGLE Media, my co-host Stu Miniman, the next guest, Joe Bida, who's the co-founder and CTO Heptio with Craig Mclucky, the famous startup that came out of the Google team, really one of the principal founders of Kubernetes with Craig and the team, Brendan Burns and the likes. Great to have you on theCUBE, thanks for coming on. Well, thank you very much for having me, it's exciting. Good time, first time on theCUBE. Bled, you've been trying to get your perspective because obviously we're fans of the Kubernetes. I just had Lou Tucker around, we've been talking inter-clouding and it's orchestration opportunity. You guys had that vision and it's really important to tell the story or at the beginning, with Kubernetes. You guys are sitting around having a little beer, free food in the Google cafeteria. I mean, what was it like? What happened? How did it all come together? All right, well, you know, it's, I started at Google, you know, probably 10, 12 years ago, did a whole bunch of stuff, eventually landed doing cloud and Craig and I started up Google Compute Engine, VM as a service. And the odd thing to recognize is that nobody who had been at Google for a long time thought that there was anything to this VM stuff, right? Because Google had been on containers for so long that was their mindset. Borg was the way that stuff was actually deployed. So, you know, my boss at the time who's now at Cloudera booted up a VM for the first time and anybody in the outside world would be like, hey, that's really cool. And his response was like, well, now what? Right, you're sitting at a prop, like that's not super interesting. How do I run my app, right? Which is, that's what everybody's been struggling with, with cloud, is not how do I get a VM up? How do I actually run my code? And as Google got more and more serious about cloud, you know, every big company wants to dog food their products. And so how do we make the experience that folks inside of Google have, developers inside of Google have, match the experience that the cloud customers have? And so the choice there was either we make everybody inside of Google start using VMs, which would have felt like that step backwards. Or we teach the rest of the world about board. Now, around the same time, Docker started getting a lot of attention and we're like, hey, those guys are onto something. They really found a good way to make this technology accessible to users sort of on a single node level. But our experience at Google really taught us that cluster of you, how do you actually create this sort of abstraction that a whole bunch of computers are one thing that you operate with? That was the thing that was going to be interesting. And so out of that, we decided Kubernetes was going to be the thing, or at least getting Borg out to the rest of the world. And we knew for it to be effective, it couldn't just be Google doing it alone. We had to do it in a way that would bring the rest of the industry with us. And so that's really the motivation behind Kubernetes. It took us about another three months to convince all the folks at Google that this was a good idea. It was controversial. The open source projects at the time were things like, the biggest things would be like Chrome and Android. Those things were, you know, the relationship with their community was very different from what we were aiming for. With Kubernetes, there are much more consumer focus versus infrastructure focus. It was early too for Google to recognize the multi-cloud world. I think, you know, some of it wasn't so much as multi-cloud as much as developers have a really strong sense of where the lock-in is where the vendor lock-in is. And we knew that if we wanted to win the hearts and minds of engineers and developers and folks that took this stuff seriously, as the underdog in the cloud world at the time, you had to really sort of go out there and build something that was going to be widely applicable. Because you don't want to invest your time and energy into something that's super specialized to one cloud. And I think, so the whole multi-cloud thing, honestly I think it's engineers and developers and operations folks that had that sense from the get go. I think we were just reacting to that. That's good instincts too. Kubernetes certainly working out today state of the union. We're still only less than three years old as a community. It seems like 20, but the momentum's been amazing. Has been a lot of revision. A lot of people have their own versions of Kubernetes, yet there's a core vanilla kind of Kubernetes. But it's working. I mean, people have gotten around this. What is the big thing that surprised you the most and where are you most excited right now where Kubernetes is at? Okay, so surprise. I mean, there's 4,100 people here at KubeCon. That's absolutely insane. I think we had this idea that it could be a thing but I don't think any of us imagined that within three years we'd be sitting here doing this type of thing. So that I think for me is the most surprising. I think that it's a challenge to take these ideas that have been successful inside of Google and translate those to the rest of the world. And it wasn't an easy or obvious thing. There was a lot of good ideas but figuring out how to get those out there. And I think that really is due to the larger community. I mean, folks like the Clinton Coleman from Red Hat coming in early with a lot of that really brought a lot of that sort of outside DNA necessary to bridge that gap. So surprising that we got here but really it took the community to make that happen. In terms of what I'm most excited about right now, I think with the announcement of EKS from Amazon it definitely feels like we're moving into a new phase of Kubernetes where folks are being much more focused on what do you do with Kubernetes versus how do you get Kubernetes running? And Kelsey tweeted at the other day but I think we've been saying it for a while. Kubernetes at its heart is a platform for building platforms. And really we viewed it from the start as a toolbox and I think we're only now starting to see what are the things are people going to be building with that toolbox? And I think that's going to be that larger ecosystem is going to be much larger than Kubernetes itself. Joe, coming into the show, there were so many announcements around Kubernetes. There's like 42 certified different versions out there. I think you can help explain a little bit because there's the big cloud guys, there's you mentioned Clayton who we had on the program earlier from Red Hat. There's all these companies that are like, oh well, Kubernetes is just like it's a piece and it's in there. Your company is around Kubernetes. So what does this mean that kind of Kubernetes is, I guess we'd say commoditized across there. I think it's a good thing for the industry but what does it mean? Why is there a need for Heptio and what do you guys see as your role in the ecosystem? Yeah, I think there's a bunch of folks that are really concentrating on how do I get Kubernetes up and running and that's one thing. And I think that landscape is going to be changing and evolving over time. We're definitely happy to help folks be successful with Kubernetes. It's one of those things we're going to do. We're doing open source projects. Service is important training with that. But when we look forward, I think a big part of it is how do we bridge the gap to integrate Kubernetes into businesses? How do we start building those next layer tools on top of it? And to some degree, it's a wild west, right? There's those 42 companies. Everybody's trying to actually find something that's going to be interesting, start solving problems. But the thing that's really encouraging to me is that there's, you know, Kubernetes is the base and we're doing work both Heptio and the community around conformance to make sure that we actually have a solid base that folks can build on top of. And then everybody is focused on how can we actually capture the attention of developers? How can we actually deliver value there? And so that's a really great dynamic. When everybody's like, I want to do something really great that people are going to get a lot out of, only good things are going to come from that. Yeah, and I liked, you know, there was a concern some people have. Last week, AWS is now all in. They've got EKS, but you had an announcement about the Heptio Authenticator, open source authentication, a little bit of a partnership with AWS that looked like even there. Maybe explain, you know, it sounds like one of the things you're building on top of this. Yeah, exactly. I, you know, like everybody else, we'd heard all the rumors about, hey, is Amazon going to do a Kubernetes offering or not? In our mind, there were two ways. Didn't they have to, Joe? Well, that's what I thought last year, but who knows, right? I think, you know, Amazon doesn't have to do anything. But when we first started Kubernetes, we reached out to the folks at Amazon, including Deepak, and we're like, hey, you guys are welcome, come join us here. And they're like, yeah, yeah, we'll join you when the customers are asking for it. Well, it turns out the customers were asking for it, and so here they are. And I think it's actually a great thing, it really is. Now, I think it could have gone two ways. They could have built in a bunch of integrations into Kubernetes that were only available through EKS, that really made EKS a sort of more integrated, better Kubernetes, then running open-source Kubernetes on top of Amazon. Or they could have worked with the community, with upstream, to try and make Kubernetes run great on Amazon, you know, better on Amazon as is, but then run even better when you're running it with EKS, and they actually have the management on top of it. And I think they decided to go that second route, which is much more community-friendly. And a couple weeks before the announcement, they reached out to us, said, hey, we noticed you had this project, it looks really interesting. We need a way to bridge IAM to authenticate two Kubernetes, and we like the approach that you're taking, can we work together to continue to develop this? And that was the first signal to us that they wanted to really reach out and work with the community, and so we're like, hey, that sounds great, let's work together and get that stuff out there. It's still very early, I think EKS is GA next year. They set an aggressive goal for themselves, and so I'm really looking forward to see where they take that, and we're going to partner with them where it makes sense around things like Authenticator. You mentioned we're going to a whole other level with Kubernetes around Amazon's announcement kind of goes to the next level. You also mentioned you worked at Google Compute, all these other coolness on Google, and you got Heptio, you're solving, making interesting things happen with Kubernetes, and you got a new class of developers coming in, they've never heard of what a local director is. So infrastructure as code is happening, so you got the cloud game going on. So I got to ask you, as Kubernetes starts to continue to take shape, a lot of people are trying to survive, right? And there's technical architecture decisions that most of tech chess game, what side of history will you be on, kind of thing going on? And customers want more clarity. So you have a lot of movement, and customers want clarity. So how do you see it continuing, and what is the right path in your mind? Because it's looking good right now, and commoditization as some say, I think is a good thing, because value is this value in interoperability, this value in orchestration, this value in a new class of web developer creating, solving problems with code, whether it's societal problems or other things. So there's a lot of big picture holistic things happening. And Kubernetes is kind of a strikes the heart of that. Yeah, what's the right path in your mind? What's the vision you think Kubernetes should go into? Well, I think first of all, I think change happens in the industry, both fast and slow, right? It feels like it's been three years since we open sourced Kubernetes, and it's come a huge way since then. That happened really fast. You look at enterprise, you look at enterprise adoption cycles. I believe last I heard the mainframe division was a growing profit center for IBM, right? This stuff doesn't go away. And so as we see things like containers in Kubernetes and serverless in cloud, as we see these things come on the scene, it doesn't necessarily replace stuff, it augments and it adds over time. So we see the mix of where people invest shift, right? So in that way, things become established quickly, but old things go away slowly, right? So I don't think it's going to be as quick of a shift as maybe it might seem at first. Now, in terms of where the opportunities are moving forward and where we see this developing, the thing that's exciting for me is as we have, and this is something early on, talking with Brendan, he got super excited about, is as we provide new abstractions, as we provide a new toolbox, how do people start creating systems and applications that take advantage of that? And I'll give you an example. Distributed systems, pre-systems like Kubernetes, were very difficult because not only did you have to do the thing that you wanted to do, you had to build all of this plumbing to actually get two things to talk to each other, to find, to secure, all that stuff had to be created from scratch. And those systems were rare and hard to manage and few and far between. Now, with things like Kubernetes, there's a whole set of problems that you actually don't have to solve. So the floor that you need, the floor is that much higher for building these systems. So I think we're going to see a shift, not just a cloud native, but I also think we're going to see a set of applications that are Kubernetes native. These are applications that assume that Kubernetes is the substrate that they're running on and they take special advantage of it. And I think we're going to see amazing things happens when we really democratize the plumbing for building distributed systems. And that's the key, make that frictionless. So if people want to go Kubernetes native, they can take advantage. Okay, that's cool. I want to get your thoughts on that. To take that to the next level, as the world of IoT comes down, you could almost look at the world now as all IoT. I mean, there's no on-prem and there's no cloud. If you believe this service mesh and unplugable architectures, you could argue that a data center is a network point. It's an attached device, it's an internet of things. So you're going to need policy. The light bulb has a process in it. The wifi has wifi's everywhere. So in a way, this is all going to be kind of a grid, if you will. It's almost going to be kind of a mesh. And this is kind of the right direction, don't you think? The more services that come online, you just want to connect to them. For sure. That's the Nirvana, right? I mean- Yeah, I think- I mean, so there's a couple of- There's a couple of- There's a bunch of trends that we're seeing happen there. I think with IoT, we see also a move towards edge computing. This idea of we're going to see much more stuff happening in a more distributed manner. Whether that edge happens to be in your house or whether it's in a telecom cabinet or whether it's just sort of mini data centers that are dropped in to parking lots here and there. So that introduces a whole bunch of new problems in terms of how do you manage that stuff at scale? And one of the things that I see is that we're seeing an interesting overlap between CDN providers and cloud providers. So you have Cloudflare introducing their cloud workers where you can start running actual code in their CDN nodes. And that's the culmination of CDN providers over time fighting with each other to provide more and more customization. On the other hand, you have Amazon taking Lambda, finding ways to actually use Lambda and push that out to the edge, even into devices that are doing local machine learning. And so there's this overlap between these two different worlds. And then also as we move stuff closer out to the clouds, the political situations that people deal with become that much more complex. As you start running compute in all these different countries, all of a sudden you can't necessarily go to one provider to actually deal with all that. So we're moving from this world where when you're centered around data, which is the traditional cloud, when you want to put it all in one big pile with compute around the edges, that's kind of like the traditional data center. Going with a few large provider makes a ton of sense. As we move towards a much more distributed world, it becomes a more distributed problem both in terms of how do you manage the compute but how do you manage the relationships and how do you actually understand what's happening across all that. And I think Kubernetes can be a part of that puzzle. For sure. But it's not the end of the answer. There's still a lot of problems to be solved there. No, no, but you get the first mile post. You can say, hey, I can start orchestrating workloads and have endpoints and have services that talk to each other. It's a good first step. Joe, one thing I wanted to ask you, what are the stumbling blocks? What do people need to look out for? Because most companies out there aren't Google. So this morning at today's keynote, you can find it online. There's that cloud native roadmap that Dan was showing. That is an interesting thing that cuts both ways. On the one hand, it shows an enormous amount of innovation. It shows that we're seeing this explosion of interest in this world, and it's really invigorating. That's from sort of an entrepreneur's view and a technologist's view. If I'm a customer, that thing's kind of horrifying, right? I look at that and I say, wow, I really have to understand all of this stuff to get ahead. And so I think the biggest stumbling block is really being able to make sense of all the noise out there. And I think that that noise is part and parcel of an active, innovative, chaotic ecosystem. But I think it's one of those things that makes it that much harder for enterprises and for sort of more mainstream developers to adopt. And so, Tim, we've been saying this for a while. For Kubernetes to be successful, we have to make it boring. Tim Hawken, I think it's probably, maybe it was the first one to say that. But we not only have to make Kubernetes boring, we have to make that entire stack boring. We have to make cloud native boring. That's one we'll have succeeded. I don't know what this console will look like when cloud native is boring, but it'll probably be very different than- It'll certainly create some excitement. Boring is reliable, boring is safe, boring is secure, boring is comfortable. I mean, Mark Zuckerberg once said, move fast, break stuff, and then he revised it to move fast and be 100% reliable. That's boring. Did he actually say that? He shifted his narrative. Because that was the maverick, you know, early days when he started running at five nines, like, oh no, the ball game. Actually, maybe it matters. Joe, great to have you on theCUBE. Thanks for sharing your awesome insight into the dynamics of the computing industry that's going cloud native, going KubeCon, and certainly Kubernetes, that you helped put together with the team, certainly taking on the life of the zone. Last minute, take a minute to talk about Heptio, what you guys are working on. Get the plug in. Yeah, so Heptio, we have services, support and training that we're offering to make customers successful with Kubernetes today. And that's been invigorating, really getting out there and talking with folks. Seeing the problems that they're hitting now versus where we want it to go. We're doing a bunch of work around open source projects. We have Heptio Arc, which is a backup disaster recovery project, open source. We have Sonnaboy, which is a diagnostic project for running the conformance tests, and it underpins the Kubernetes conformance effort. We have Ksonnet, which helps you configure applications. And then we also have Contour, which is an ingress controller building on top of Envoy, another CNCF project. And then into 2018, we're going to be offering more products and projects and services that really start targeting the special needs of larger and larger enterprises. And that's where our focus is going to shift over time. And you guys are certainly helping customers who are under pressure to add more services. I mean, look at what Amazon's doing, more announcements. I mean, they're all little announcements, mostly some of them big, some little, but still, the cadence of new things happening is fast at all times right now. I can't keep up either, nobody else can. We try, two and a half hour keynotes, ridiculous. Joe Beta here inside theCUBE, co-founder, CT of Heptio, hot startup, making Kubernetes interesting and exciting, and reliable and boring, not boring, should say that, but interest research is good. It's theCUBE bringing you all the live action from Austin, Texas. I'm John Furrier, Stu Miniman, KubeCon and CloudNativeCon. We'll be right back after this short break.