 Live from Copenhagen, Denmark. It's theCUBE, covering KubeCon and CloudNativeCon Europe 2018. Brought to you by the CloudNative Computing Foundation and its ecosystem partners. Welcome back everyone. Live here in Copenhagen, Denmark, it's theCUBE's coverage of KubeCon 2018. I'm John Furrier, the host of theCUBE, along with Lauren Cooney, who's the founder of Spark Labs. She's been co-hosting for two days, two days of wall-to-wall coverage. Step on fabled products and strategy lead at Canonical is here inside theCUBE and from San Francisco. Again, welcome to theCUBE. Thanks for coming. Thanks so much for having me. So I got a, you guys have been around the block. You know about open source software platforms. You know, you've been doing it for a while. Interesting time here at KubeCon. Kubernetes, Istio, Kubeflow, CloudNative. They still got the branding CloudNativeCon and KubeCon. Modern application architectures now in play. You see this notion of an interoperability model coming in that's certainly going to be a de facto standard. People are already kind of declaring it a de facto standard. Really shows a path to multicloud, but also frees up developers from a lot of the heavy lifting. Lou Tucker from Cisco is saying, they don't want to do networking. Let's just have that be infrastructure as code. That's DevOps. That's what we want. That is exactly right. What are you guys doing here? What's the story with canonical and how does that fit into the megatrend? Yeah, I mean, there's a couple of things that we've, we at canonical always believed to be one of the core sort of tenants in our distribution of Kubernetes. As you know, we've been very active in the space fairly early on and have been an active distributor of Kubernetes and a certified distributor of our, you know, of our version of Kubernetes. Pure upstream remain conformant to the main public clouds such as to enable that workload migration and mobility from on-prem up to any of the other providers to accommodate all kinds of use cases. You guys made a bet on Kubernetes. Obviously, good call, right? So what's the progress now? What's next? Because the bets are paying off. So Red Hat had a great bet with what they did with Kubernetes, changed what OpenShift became. You guys had a bet in Kubernetes. What has that become for canonical? Yeah, so based on the pure upstream distribution that we have, we really feel that enabling the ecosystem in a standards compliant way so that all of the landscape projects that are part of the CNCF can be deployed on top of Kubernetes, on top of our distribution of Kubernetes in just the same way that they would be developed or deployed in any of the large containers of service offerings that are out there is one of the big benefits that our customers would gain from using our... And what's your differentiated for the distribution of Kubernetes you have versus others? The first one, I think, is the notion that deploying Kubernetes on-premise is something that you want to do in a repeatable fashion and operationally efficient with the right CAPEX-OPEX mix. So we believe that there is a place for Kubernetes as a product, just deploy it, works on any substrate that you've got available to you. But then also, for Main Street America, right? You may want to have a managed service on top of Kubernetes as well. And we offer that too, just a way to get started and kick the tires and see whether that takes you as far as the developers are concerned. Now, on-prem, you will find that there are a couple of challenges when deploying Kubernetes that are really the key differentiator. The first one, I would say, is things like integration into the storage, that's local, integration into the network, that's local, and integration into all of those services that should be available in a cloud-native, microservices architecture platform, such as low-balancers, right? Elasticity, object store, et cetera. The second, and most importantly, because it is a key enabler for those next-generation workloads, is the GPGPU enablement work that we're doing with partners such as NVIDIA. So when you deploy the canonical distribution of Kubernetes, you actually get the NVIDIA acceleration out of the box the way that NVIDIA envisions this on top of Kubernetes, and the way that it is, by the way, being deployed on the public cloud. So you bring a lot of your goodness to the table inside the Kubernetes distribution. Okay, what are some customers doing? Give some use cases of some customers, Kubernetes, what are some of the things that they're doing with it? What's the early indication? What's the feedback? Sure, so we have a ton of customers that are using our version of Kubernetes to do the machine learning applications and the AI of the next-gen workloads in use cases such as smart cities or connected cars, where when you look at self-driving cars, right, as the next-gen that's coming out of the valley, you know, they put in 300,000, 150,000, 400,000 miles a year on the road these days just optimizing the models that are being used to actually take over one day. And so enabling those kind of workloads in a distributed fashion requires DevOps expertise. Now, the people who are actually writing those applications are not DevOps people, they're data scientists, right? They shouldn't have to learn how to deploy Kubernetes, how to create a container and all those things. They should just be able to deploy the application on top and the attractive substrate that actually supports that distributed application use case. And so that is where we come in. This is interesting because what you're basically doing is making an application developer, a DevOps developer overnight. That's exactly right. And that's really important. I was just talking with the co-chair of CNCF, we're talking about Liz Rice and I were talking about why everyone's so excited here. And one of the things I said was because people who were doing DevOps were hardcore and they had to build everything from scratch and all those scar tissue. But the benefits, once you got through the knot hole there, the benefits were amazing, right? So you go, okay, you don't want to do that again, but now there's a way to make it easier. So there's kind of a shared experience, even though no one has met each other, there's kind of a joint community. I agree. So I think it is about increasingly about enabling developers who are experts in their field to actually leverage Kubernetes and the advantages that it brings in a more intuitive fashion. So just take it up a notch. How did the Kubernetes vibe integrate in with Canonical? I'm sure, given the background of the company, it probably was a nice fit. People embraced it. You guys were early. Yeah. I mean, what's the internal scuttle butt on the vibe of Kubernetes? Oh, we love Kubernetes as a technology. The Ubuntu was always close to the developer and close to where the innovation happens. So it was a natural fit to actually support all that workflow now in this new world of Kubernetes. We embraced OpenStack for the same reason. And in a similar fashion, Kubernetes has really driven the point home. Containerist applications with a powerful orchestration framework such as Kubernetes are the next step for all the developers that are out there. And so as a consequence, this was a perfect match. You know, and it's also a no-brainer if you think about it. You look at software methodology moving to the next level. This is total step-up function for productivity for developers. Yeah, yeah. And that's really a key thing. What's your observation of that trend? Because at the end of the day, there's now Kubernetes, which does a lot of great things. But one of the hottest areas is Istio, service mesh, and then you got Kubeflow orchestration, a lot of other things that are happening around Kubernetes. What are you guys seeing that's important for Canonical's customers? What you're doing product-wise? Where's the order of operations? What's next? What are you guys focusing on? What's the priorities? Well, our biggest priority right now is enabling things like Kubeflow, which by the way are also using Istio internally, right? To actually enable those data scientists who actually deploy their I workload. So we worked very closely with Google to try and enable this in an on-prem fashion out of the box, which is something you can actually do today. So you guys are doing this now inside Canonical? We're doing this right now. And this is also where we're going to double and triple down. Well, and this is actually a best practice, too, if you think about it. You want to take it in-house and then get a feel for it. So what's the internal vibe on that? Positive? Oh, absolutely. We always saw infrastructure as code and actually as intelligent infrastructure as something that we wanted to build our conceptual framework around. So very concretely, right? We've always had this notion of, composable building blocks adding up to some of two, some of one being greater than two, right? Like those types of scenarios. And so actually using things like Kubernetes as an effective building block to then build out web applications that use things like machine learning algorithms underneath, that's a perfect use case for our next gen workload and also something that we might use ourselves internally. Well, hey, that whole building block thing, it's happening, news flash. Exactly, right? I mean, it's almost a pinch me moment for the people in the industry, like, oh my God, it's going to go to a whole new level. How do you guys envision that next level going? And beyond the building blocks, how is it, I mean, what's the vision that you guys have? Obviously infrastructure as code, programmability, but now you're talking about infrastructure as code was great, but now you've got microservices growth coming on top of it's a services market now. It is, it is. I think that the biggest challenge will be the distribution of the workloads, right? So you have edge compute coming along in the telco space. You have, like I said, smart cities, right? The sensors will be everywhere and they will feed data back and how do you manage that at scale, right? How do you manage that across various different hardware perspectives? We have hardware platforms such as ARM64 picking up, right, and actually playing a very significant role at the edge and increasingly even in the core. And so we've always believed that providing that software and the distribution of IS such as Kubernetes and others on top of those additional architectures would make a huge difference and that is clearly paying off. So what we see is the increased need of managing hybrid workloads across multi-cloud scenarios that could be composed of different architectures, not just x86, the future is not homogeneous at all. It'll be all over the place and all those use cases and all those particular situation require that building block principle like all the way from the OS up to the application. And that's a great use case for containers, Kubernetes, Istio, Kubeflow, all stacking in line beautifully from an evolution standpoint. We're going to ask you a personal question. I mean, I was a canonical, great company. I want to thank canonical for being a sponsor of the Kube over the years. We've had Mark Shuddleworth on the Kube at an open stack going way back when you guys are a great participant in the community as a company and the people there have been phenomenal. You're new. I'm new. What attracted you to canonical? What was the motivating force? What drew you in? You're now running products, a big job, got a lot in front of you. Obviously it's a great market, so a great company. Just share, just color and why, why canonical, what attracted you there? You know, I've always been a user of Ubuntu. I've been a user since the first hour. I've used Ubuntu in my research. I did robotics based on Ubuntu way before it was cool. I built all kinds of things on top of Ubuntu throughout my entire career. And so working for canonical, which is a company that always exhibited great vision into the future and great predictions into trends that would prove to become true, was just for me, something that was very attractive. Yeah, their leadership has a good eye on the prize and they had good 20-mile stairs, we say. They can see the roadmap ahead and then either course corrections or tweaks. Great, awesome. Well, I mean, what's new there? I mean, let's take a minute to explain what's new at canonical. Role here at KubeCon, what are some of the conversations you're having? Yeah, so I mean, for us at KubeCon, it's always been an important part of our outreach to the community, great opportunity for us to have great conversations with our partners in the field. I think it is really about enabling the ecosystem in a more straightforward way. There's no better place to have those types of conversations than here, where everybody comes together and really establishes those relationships. So for us, it is about, again, enabling the developer and really staying close to that innovation and supporting that in an optimal way. So, yes, I mean, that to us is the role that- And you've got a lot of end users here who are building stuff. Oh, absolutely, yeah. And they, I mean, I had a talk today about Kubeflow with Google and after the talk, lots of folks came up to me and said, hey, how can I use this at home? Sometimes, whether it's timing, technology, all the above, Kubernetes really hit it strong with the timing. Industry was ready for it. Containers had a nice gestation period. People know about containers. Absolutely. The engineers know containers, know about those kinds of concepts. Now we're at a whole nother operating environment. Absolutely. You guys are at the forefront. Thanks for coming on theCUBE. Oh, thank you. I appreciate it. I appreciate sharing the perspective of 7Fable. Running product and strategy for Canonical. Building stuff. This is what's going on in Kubernetes and KubeCon and users are actually building and orchestrating workloads. Multi-Cloud is what people are talking about and the tech to make it happen is here. I'm John Furrier with theCUBE. Stay with us for more live coverage here at KubeCon 2018. Part of the CNCF Kube coverage. We'll be right back after this short break.