 Live from Seattle, Washington, it's theCUBE covering KubeCon and CloudNativeCon North America 2018. Brought to you by Red Hat, the CloudNative Computing Foundation and its ecosystem partners. Hey, welcome back everyone. We are here live with Kube coverage at KubeCon, CloudNativeCon 2018 in Seattle. I'm John Furrier with Stu Miniman, your host. All week, three days of coverage. We're in day two, 8,000 attendees, up from 4,000, expanding to China, in Europe, everywhere, the CNCF is expanding, the Linux Foundation and the ecosystem is expanding. We're here, Dan Kahn is the executive director of the CNCF. Dan, great to see you. I know you work hard. I see you out in China. You know, you've done the work, you guys and the team have taken this hockey stick, as it's described on the Twitter sphere, really up and to the right, you doubled. It's almost like Moore's Law for attendees, doubling every six months. It's really a testament of how it's structured, how you guys are managing it, the balances that you go through. So congratulations. So thank you very much and I'm thrilled that you guys have been with us through that whole ride that we met here in Seattle two years ago at the first KubeCon we ran with 1,000 attendees and here we are eight times higher two years later. But I absolutely do need to say, it is the community that's growing and we try and organize them a little bit and harness some of that excitement and energy and then there is a ton of logistics and effort that it takes to go from 28 members to 349 to put on an event like this. But we do have an amazing team at the Linux Foundation and this is absolutely an all hands on deck where the entire events team is out here and working really hard. You guys are smart, you know what you're doing and you have the right tone and posture but you set it up right. So it's end user-driven, it's open source community is the core of the event. You're seeing end users that have contributed, they're now consuming, you have vendors coming in but you set the nice playbook up and the downstream benefits of that open source core has impacted IT, developers, average developers and this is the magic. How, and you guys don't take too many hard stands on things, you let, you take a good enough stand on the enablement piece of it. This is a critical piece, explain the rationale because I think this is a success formula. You don't go too far and say, here's the CNCF stack. You pull back a little bit on that and let the ecosystem enable it. Talk about that rationale, I think this is an important point. Sure, and I would say that one of the huge advantages that CNCF has had is that we came later after a lot of other projects. So our parent, the Linux Foundation has been around for 15 years. We've been able to leverage all their expertise. We looked at some of the mistakes that OpenStack and Apache and ITF and other giants who came before us did. Our aspiration has always been to make entirely new mistakes rather than to replicate the old ones. But as you mentioned, the end user is a key focus. So when you look at our community, how CNCF is set up, we have a governing board that's mainly vendors. It does have developer and other reps on it. We have our technical oversight committee of these nine experts, kind of like our Supreme Court. And then we have this end user community that is feeding requirements and feedback back to the other group. I want to ask you about the structure. I think this is important because you guys have a great governance model, but you have this concept of graduation. You have Kubernetes and it's really solid. People are very happy with it. And there's always debates in open source, as you know, but there's a concept of graduate. I think we're going to have projects that explain that dynamic. I've heard people say, oh, it's part of the CNCF. Well, it hasn't graduated, but it's a project. It's important. It's a laddering there. Explain that concept. I think this is important for people to understand that you're open, but there's kind of a model of graduation. What does it mean? Sure. And people have said, oh, you mean they've graduated so they've left now, right? Like the kids leaving home. And it's definitely not that model. Kubernetes is still very much part of CNCF. We're happy to do it. But we think that one of CNCF's functions is as a signaling and a marketing to enterprise users. And we like the cliche of crossing the chasm where we talk about 2018 was really the year that Kubernetes crossed the chasm, went from those early adopters who've been using it for years and were thrilled with it, but they actually jump over now to the early majority. I will say though that the late majority, the laggards, the skeptics, they're not using these technologies left. We still have a ton of opportunity for years to come on that. So we say the graduated projects, which today is not just Kubernetes, but also Prometheus and Envoy, those are the ones that are suitable for really any enterprise company and that they should feel confident. These are very mature, serious technologies for companies of all size. The majority of our projects are incubating. Those are great projects, technically capable. Companies should absolutely use them if the use case fits, but they're less mature. And then we have this other category of the sandbox, 11 projects in there. And we say, look, these are incredibly promising. If you are technical enough and you have the use case, you absolutely should consider it. But they are less mature. And then our hope is to help the projects move along that graduation. And that's how companies start. Bloomberg explained that. They jump in the sandbox. They'll start getting some code in there that'll attract some people. They get their code. They don't have to come back after the fact and join in. So you get the sandbox, you got projects, you got graduation. Now Bloomberg's a little bit unusual and I like them as an example where they have, I don't know if they mentioned this, but almost a philosophy, not to spend money on software. And of course that's great. All of our projects are free and open source and they're willing to spend money on people and they hire a spectacular group of engineers and then they support everything in house. But in reality, the vast majority of end users are very happy to work with a vendor including a lot of our members and pay for some of that support. And so Bloomberg can be a little bit more adventurous than many I think. Yeah. Dan, I wonder if you can provide a little bit of context. I hear some people look at really kind of the conformance and certification that the CNCF does. And I think in many ways, learn from the mistakes of some of the things we've done in the past. Because they'll see, there's so many companies, it's like, well, there's too many distributions and maybe it could help explain the difference between a distribution and what's supported and how that makes sense. When you look back at, and we just had, CNCF just had our three year birthday this week. We have a little birthday cake on Twitter and everything. But if you look at all the activities we've been involved in over those three years. KubeCon, CloudNativeCon, we have a service provider program, we've done a lot of marketing, helping projects. I think it's the certification and the software conformance is the single thing that we've had done that's had the biggest impact on the community. And the idea here is that we wanted a way for individual companies to be able to make changes to Kubernetes because they all want to, but to still have confidence that you could take the same workload and move it between the different public clouds, between the different enterprise distros, or just vanilla Kubernetes that you download or different installers out there. And so the solution was an open source software conformance project that anyone can download these tests and run them. And then a process where people upload the test results and say, yes, my implementation is still conformant. I've made these changes, but I haven't broken anything. And we really have some amazing cases of our members, some of our biggest members, who had turned off APIs, maybe in their public cloud, for good reasons. They said, oh, this doesn't apply or we don't, but that's exactly the kind of thing that can cause incompatibility. Yeah, I mean, that's critically important. The other thing is what I haven't heard is there's so many projects here. And we go to the Amazon show and it's like, I'm overwhelmed and I don't know what to do and I can't keep up with everything. I'm actually surprised I don't hear that here because there are pockets and this is multiple communities, not like a single monolithic community. So you've got, you know, Envoy has their own little separate show and operators has a thing on Friday that they're doing and there's the helm community and sometimes I'm putting many of the pieces together but oftentimes I'm taking, you know, just a couple of the pieces. How do you, you know, manage this, you know, loosely coupled? It's like a distributed architecture. Loosely coupled is a key phrase. I think the big advantage we have is our anchor tenant of Kubernetes has its own gravitational field. And so from a compatibility standpoint, we have this, excuse me, certification program for Kubernetes and then all of the other projects essentially ensure they're kind of orbiting around and they ensure that they're compatible with Kubernetes. That also ensures they're compatible with each other. Now, it's definitely the case that our projects are used beyond just Kubernetes. We were thrilled with Amazon's announcement two weeks ago of commercial support for Envoy and talking about how one of the things they loved about Envoy is that it doesn't just work on Kubernetes, they can use it on their proprietary ECS platform on their regular EC2 environment as well. And that's true for almost all of our projects. Prometheus is used in Mezos, is used in Docker Swarm, is used in VMs. But I do think that having so much traction and momentum around Kubernetes just is a forcing function for the whole community to come together and stay compatible. Well, you guys did a great job. That happened last year. It really to me is an example of a historic moment in the computer industry because this is a modern version of an enabling technology that's going to enable a lot of value creation, a lot of wealth creation, a lot of customer, and it's all in a new way. So I think you guys really crack the code on that and it will continue success. You kind of say China going gangbusters, you're expanding. China, by the way, is one of the largest areas we've reported on SiliconANGLE.com on theCUBE in the past. China has emerged as one of the largest contributors and consumers of open source, given the rise of all the action going on in China. Huge. We've been thrilled to see that. And I mean, there was just the example yesterday where SCD is now the newest project, the newest incubating project in CNCF and the co-creator of that and really the lead maintainer for it. Left CoreOS when it was acquired by Red Hat and is now with Alibaba. And he's originally from China. He's helping Alibaba just, who's a Platinum member of CNCF, who's been offering a certified Kubernetes service, but they're now looking at how they can move much more of their internal workloads over to it. JD.com has 25,000 servers. That's the second biggest retailer in China. So I was there six times last year. I know I ran into you once in a hotel lobby. But- What are you doing in China? It's huge. We're here. No, this is a big dynamic. This is new. I mean, this is a big force and function. And to have so much energy, and I do also want to really emphasize the two-way street that it's not just Chinese companies adopting these technologies that started in the US. They're contributing. We were thrilled a month ago to have Harbor come in as an incubating project. And that started in China and is now being used across the world. So Dan, 2019, you've got three shows again. Barcelona, Shanghai, and San Diego. Exactly. Of course the numbers are going to be up in the right, but what else would we be looking for? So I think the two, so definitely China, we're going to continue doing it there. We continue to be really interested in serverless. We're thrilled with the progress of our serverless working group. They have this new cloud events spec. We have all of the different major clouds participating in it. The third area that I think you're going to see us that is somewhat new is looking at telcos. And our vision is that you can take the most networking code today is done in virtual machines called virtual network functions. We think those should evolve to become cloud native network functions, that same networking code running in containers on Kubernetes. And so this is actually going to be our first time with a booth at Mobile World Congress in Barcelona in February. And we're going to be talking about it. Makes a lot of sense. IoT over the top, a lot of enablement there, eliminates and makes some efficiencies in that inefficient stacks. Yeah, and on the edge as well. Yeah, thanks for coming on. I appreciate it. And again, you've done the work hard work to continue great success. Congratulations. I know it's early days still, but. I hope it is. I don't, it's some date Kubernetes is going to plateau, but really doesn't feel like it'll be 2019. Yeah, it definitely is not boring even though it has matured now. Dave Collins, executive director of the CNCF here on the side of the queue breaking it all down. Again, another successful show. Just the growth. This is the tsunami. It's the rise of Kubernetes and the ecosystem around it creating values. Cube coverage live here in Seattle. We'll be back with more coverage after this short break. I'm John Furrier with Stu Miniman. Be right back.