 Welcome back everyone to theCUBECon. Cloud NativeCon coverage, I'm John Furrier, host of theCUBE. We're here in person, 2020, 2020, a real event. It's a hybrid event. We're streaming live to you with all the great coverage and guests coming on. Next three days, Clayton Coleman's here, Chief Hybrid Cloud Architect for Red Hat. He's joining me here to give our overviewers talk but also talk about hybrid cloud, multi-cloud, where it's all going, where Red Hat's doing. Great to see you. Thanks for coming on. It's a pleasure to be back. It's a pleasure to be back at theCUBECon. It's an honor to have you on as a Chief Architect at Red Hat on Hybrid Cloud. It is the hottest area in the market right now. The biggest story, obviously we're back in person, that's the biggest story here. The second biggest story that's been the most important story is hybrid cloud. And what does it mean for multi-cloud? This is a key trend. You just gave a talk here. What's your take on it? You know, I like to summarize hybrid cloud as the answer to, or it's really the summarization of yes, please, more of everything, which is we don't have one of anything. Nobody who's got any kind of real footprint is single cloud. They're not single framework. They're not single language. They're not single application server. They're not single container platform. They're not single VM technology. And so, and then, you know, looking around here in this, you know, partner space where eight years into Kubernetes and there was an enormous ecosystem of tools, technologies, capabilities, add-ons, plugins, components that make our applications better. The modern application landscape is so huge that I think that's what hybrid really is, we've got all these places to run stuff more than ever and we've got all this stuff to run more than ever and it doesn't slow down. So how do we bring sanity to that? How do we understand it, bring it together? And Kubernetes has been a big part of that. Like it unlocked some of that. What's the next step? Yeah, that's a great, great commentary. I wanted to take into the Kubernetes piece, but you know, as we've been reporting the digital transformation at an all-time high, speed is the number one request. People want to go faster, not just speeds and fees, but like ship code faster, build apps faster, make it all run faster and secure. Okay, check, get that. Look what we were 15 years ago, 10 years ago, five years ago, 2016, the first KubeCon in Seattle, we were there, the small event, Kubernetes, we got to sell it, figure it out. Convince people that it's worth it. Yeah, so what's your take on that? I mean, obviously it's mature, it's kind of de facto standard at this point. What's missing, where is it? So I think Kubernetes has succeeded at the core mission, which is helping us stop worrying about all the problems that we spent endless amounts of time arguing about. How do I deploy software? How do I roll it out? But in the meantime, we've added more types of software, you know, the rise of AI ML, you know, the whole ecosystem around training software models. Like what is an AI model? Does it look like an application? Does it look like a job? It's part batch, part service. It's spread out to the edge. We've added mobile devices, the explosion of mobile computing over the last 10 years has co-evolved. And so Kubernetes succeeded at that kind of set a floor for what everybody thought was an application. And in the meantime, we've added all these other parts of the application. It's funny, you know, Dave Vellante, we're talking about, and it's Stu Miniman, who now works at Red Hat, who'll be on later. Back in the first two cube cunts, we're like, you know, this is like a TCP-IP moment in the OSI model. That was a killer part of the stack. Now, it was all standardized below TCP-IP. Kubernetes feels like a similar kind of construct where it's unifying, it's creating some enablement, it's enabling some innovation. And it kind of brought everyone together. At the same time, everyone realized that it's real. The whole cloud native is real. And now we're in an era now where people are talking about doing things that are completely different. You mentioned, is it a batch? Is it a job? How is AI? New software paradigm, development paradigms, not the software development lifecycle, it's just like software development in general is impacted. Absolutely, and the components, like we spent a lot of time talking about how to test and build application, but those are things that we all kind of internalize now. We have CI CD processes. CI CD is critical because it's going to be in lots of places. People are looking to standardize, but sometimes the new technology comes up alongside the thing we're trying to standardize and we're like, well, let's just use the new technology instead. Function as a service is kind of a, it came up, Kubernetes grew K native and then you see the proliferation of functions as a service choices. What do people use? So there's a lot of choice and we're all building on those common layers, but everybody kind of has their own opinions. Everybody's doing something subtly different. Let me ask you your opinion on more under the hood kind of complexity challenge. There's general consensus in the industry that there's a lot of complexity. Okay. You don't need to debate that. But that's in a way a good thing. And since if you solve that, that's where innovation comes in. So the goal is to solve complexity, abstract out of the heavy lifting, understeer heavy lifting, as Andy Jackson and I would say, or abstract away complexity and make things easier to use. Well, and open source and this ecosystem is an amazing, it's one of the most effective methods we've ever found for trying every possible solution and keeping the five or six most successful. And that's a little bit like developers, developers flow downhill. Developers are gonna do what's easy. If it's easier to put a credit card in and go to the public cloud, you're gonna do it. If you can take control away from the teams at your organization that are there to protect you, but maybe aren't as responsive as you like, people will go around those. And so I think a little bit of what we're trying to do is what are the commonalities that we could pick out of this ecosystem that everybody agrees on and make those the downhill path that people follow? Not putting a credit card into a cloud, but offering a way for you not to think about what cloud you're on until you need to, right? Cause you want to go to the fridge as a developer, you want to go to the fridge, pull out your favorite brand of soda. That favorite brand of soda might have an AWS label on it. So talk about the, the open shift and the Kubernetes relationship and how you guys pushed that boundaries. Dennis being control plane and nodes. These are things that you talked about in your talk. Talk about, cause you guys made some good bets on open shift. We've been covering that. How's that playing out? And what's the relationship now? It is interesting, coming into Kubernetes, we came in from the platform as a service angle, right? Platform as a service was the first iteration of trying to make the lowest cost path for developers to flow to business value. And so we added things on top of Kubernetes. We knew that Kubernetes would be complex. So we built in a little bit in our structure, in our way of thinking about cube that it was never going to be just that basic bare bones package that you were going to have to make choices for people that made sense. Obviously as the ecosystem's grown, we've tried to grow with it. We've tried to be a layer above Kubernetes. We've tried to be a layer in between Kubernetes. We've tried to be a layer underneath Kubernetes. And all of these are valid places to be. I think that next step is we're all kind of asking, we've got all this stuff. Are there any ways that we can be more efficient? So I like to think about practical benefits. What is a practical benefit that a little bit of opinionation could bring to this ecosystem? And I think it's around applications. It's being application centric. It's what does a team 90% of the time need to be successful? They need a way to get their code out. They need to get it to the places that they want it to be. And that place is everywhere. It's not one cloud or on-premise or a data center. It's the edge. It's running as a lambda. It's running inside devices that might be coming, being designed in this very room today. It's interesting, you know, you're an architect, but you know, obviously the computer science industry is the people that were trained in the area are learning. It's pretty fascinating and almost intoxicating right now in this market because you have an operating system dynamic, a systems kind of programming model with distributed cloud edge is on fire. That's only going to get more complicated with 5G and high density data applications. And then you've got this changing modal mode of operations for programming with bots and AI and machine learning, two new things, but it's kind of the same distributed computing paradigm. What's your reaction to that? Well, and it's interesting. I was, I kind of described like layers we've gone from Linux replaced proprietary UNIX or mainframe to virtualization, which then we had a lot of Linux. We had some windows too. And then we moved to public cloud and private cloud. We brought config management in, we moved to Kubernetes. We still got that OS at the heart of what we do. We've got application libraries and we've shared services and common services. I think it's interesting like to learn from Linux's lesson which is we want to build an open expansive ecosystem. Yeah, here. Kind of like, kind of like what's going on. We want to pick enough opinionation that it just works because I think just works is what, let's be honest, we could come up with all the great theories of what the right way computers should be done but it's going to be what's easy, what gets people, help them get their jobs done. Trying to take that from where people are today on Qube, in cloud, on multiple clouds, give them just a little bit more consolidation. And I think it's a trick people or convince people by showing them how much easier it could be. Yeah, what's interesting around what you guys have done at Red Hat is that you guys have real customers that are demanding and you have enterprise customers. So you have, you're on the front edge of the bleeding edge making things easier. And I think this good enough is a good angle but let's face it, people are just lifting and shifting to the cloud now. They haven't yet refactored. And refactoring is a concept of taking what you're doing in the cloud and taking advantage of new services to change the operating dynamic and value proposition of say the application. So the smart money is all going there. You're seeing the funding come into applications that are leveraging the new platform, replatforming and then refactoring. What's your take on that? Cause you got the edge and you got other things happening. There are so many more types of applications today. And it's interesting because almost all of them, start with real practical problems that enterprises or growing tech companies or companies that aren't tech companies but have a very strong tech component, right? That's the biggest transformation in the last 15 years is that you can be a tech company without ever calling yourself a tech company because you have a website and you have an app and your entire business model flows like that. So there is, I think pragmatically people are, they're okay with their footprint where it is. They're looking to consolidate. They're very interested in taking advantage of the scale that modern cloud offers them. And they're trying to figure out how to bring all the advantages that they have in these modern technologies to these new footprints and these new form factors that they're trying to fit into whether that's an application running on the edge next to their load balancer in a gateway in Telco. 5G is happening right now. Red Hat's been really heavily involved in the Telco ecosystem. And it's Kubernetes through and through. It's building on those kinds of principles. What are the concepts that help make a hybrid application, an application that spans the data flowing from a device back to the cloud, out to a gateway, processed by a big data system in a private region, someplace where compute is cheap? Can't be a silo. No, absolutely not. It has to be distributed non-siloed based model. And how do we do that and keep security? How do we help you track where your data is and who's talking to whom? There's a lot of people here today who are helping people connect. I think that next step, that connectivity, the knowing who's talking and how they're connecting, that'll be a fundamental part of what emerges as the next door. I think the observability to me is the data. It's really about a data funding, a new data sector, the market that's going to be addressable. Having data addressability is critical. Clayton, I really appreciate you coming on and giving your perspective and expert in the field. I got to ask you, I got to say from a personal standpoint, how open source has truly been a real enabler. You look at how fast new things could come in and be adopted and vetted and things get kicked around. People try stuff that fails, but they build on each other, right? So AI, for example, is just a great example of look at what machine learning and AI is going on. How fast has been adopted? Absolutely. I think that would have been done on open source. I have to ask you guys at Red Hat, as you continue your mission and with IBM with that partnership, how do you see people participating with you guys? Obviously you're here, you're part of the ecosystem, big player, how are you guys continuing to work with the community? Take a minute to share what you're working on. So first off, it's impossible to get anything done, I think in this ecosystem without being open first. And that's something that Red Hat and IBM are both committed to. A lot of what I try to do is I try to map from the very complex problems that people bring to us, because every problem in applications is complex at some layer. And you've got to have the expertise, but there's so much expertise, so you've got to be able to blend the experts in a particular technology, the experts in a particular problem domain, like the folks who consult or contract or help design some of these architectures or have that experience at large companies and then move on to advise others in how to proceed. And then you have to be able to take those lessons, put them in technology and the technology has to go back and take that feedback. I would say my primary goal is to come to these sorts of events and to share what everyone is facing, because if we as a group aren't all working at some level, there won't be the ability of those organizations to react, because none of us know the whole stack. None of us know the whole set of details. And the stack's changing too. I mean, you got, again, I've referenced OSI miles more of an eighties metaphor, but that changed the game on proprietary and that was the beginning. It allows us to think and to separate. You want to have nice thin layers that the world on top doesn't worry about below, except when you need to and below you can make things more efficient and public cloud, open source, Kubernetes and the proliferation of applications on top, that's happening today. I mean, Paul Moritz used to talk about the hard and top when he was the VMware CEO back in 2010. I remember him saying that. He essentially predicted the whole, we call it the mainframe in the cloud at the time because it was a funny thing to say, but it was really a computer. I mean, essentially distributed nature of the cloud. It happened. Absolutely. Clayton, thanks for coming on theCUBE and sharing your insights. Appreciate it. It was a pleasure, thank you. All right, Clayton's here on theCUBE. I'm John Furrier, here live in LA for KubeCon, CloudNativeCon in person. It's a hybrid event, we're streaming, also going to the CUBE platform as well. Check us out there, all the interviews. Three days of coverage, we'll be right back.