 Live from San Francisco, it's theCUBE. Covering Red Hat Summit 2018. Brought to you by Red Hat. Hey, welcome back everyone. We're here live in San Francisco, Red Hat Summit 2018's CUBE's exclusive coverage. We're out in the open in the middle of the floor here, as open source is always done out in the open. This is theCUBE doing our part, extracting the simulators. I'm John Furrier, the co-host of theCUBE with John Troyer, my co-host analyst this week. He's the co-founder of a firm, advisory firm. Arganesca is Paul Comey, a president and products and technology at Red Hat. Architecting the future of Red Hat and products and technologies all open source. Great to see you again. Great to see you. So thanks for coming on. So great keynote today. You guys have done a great job here. I thought the messaging was great, but the excitement was strong. We just came back off of a week in Copenhagen, CUBECon, where Kubernetes clearly sees the de facto standard around Kubernetes, the core of Kubernetes, with a lot of room to differentiate around. You got Istio, service meshes, a lot of exciting things for application developers. And then under the hood, there's new life being brought into OpenStack. So there's clear visibility now into what's going on, swim lanes, whatever you want to call it. People kind of see it. So congratulations. Thank you. Magical moment, lucky strike, all on the cards. Give us some color. You guys been working on this for a while. Go back and we're going to all start. And when did things start clicking together for you guys? Well, I know I sometimes sound like a broken record here, but I mean, the key to our success is the commercialization of Linux. I mean, you know, Linux, we started Linux as a commodity play. You know, it was cheaper, cheaper, almost as good, et cetera, but it became such a powerful platform. All the innovation you just talked about is built around Linux. It's all tied into Linux. So once we laid down the Linux base in the customer data centers, which is such the logical extension to go to these new technologies because you really need to be a Linux vendor in order to be able to do a Kubernetes release, to be able to support a containers release, any of these things, it's all just intertwined into Linux. And your model is working, obviously, the open source is no secret that that's open wins over proprietary and closed. But you guys also have a community model that's feeding into the price of technology. Jim White has went into detail on, hey, you have a crystal ball in technology because you're smart guys, but ultimately the users in the communities give you direct feedback of what's relevant and cool at the right time. This is really where Kubernetes, lucky strike for you guys was really there. You saw it, so the commitment you jumped in. Can you explain that dynamic of how the products get fed in from the communities? I'll give you actually a better example of OpenShift itself. So we originally started OpenShift back in 2011 and we started it as a marketing project. We started it as a cloud-based platform to get developers out there building to our platform. And a lot of our customer-based saw it and came to us and said, I want this as a product. This is really, really powerful. So we made a product out of it. First one, Kubernetes wasn't around, containers weren't around, we built it on virtual machines. We had what we called gears to lock in. And then containers started to morph in and by release three, we transformed it to containers. Then we brought in Kubernetes because we had worked with the Google folks earlier on that. So we really listened to our customers. We started at something we thought was going to be an expense. And it turns out to be one of our hottest platform right now based on what our customers in the community told us. Timing's everything too. And the good timing is as the clouds scale started also becoming relevant. You see Amazon's success. Now you got Azure, IBM and everyone's kind of seeing that opportunity. How are you guys looking at the container piece? Because we can look at the history and Docker, trying to monetize too early. We've documented that well in the queue many times. CoreOS, a recent acquisition, big one for you guys and strategic but also a great team. Containers are super important. Talk about the role of container specifically. Not so much as a business model, but as a linchpin between how orchestration is moving and how these service meshes are coming out. I mean, think just real quickly what containers are first. Containers are just Linux carved up in a different way. Still have a kernel, still have user space. The difference is you take just the user space you want with the application and you run it that way. So all the same life cycles, security issues you have to fix, et cetera. You have to do in a standard Linux distro you have to do it in containers. The first thing, containers have been around forever. They were in Unix, if we all remember. But the killer app for containers was because now when I can bring just enough of the OS with the application, I can run that to the cloud. That's how we get the app out to the cloud. That's how we get it onto the private cloud, out to any of the public clouds, how we reverse the clouds. So even though they've been around for a while, it's the killer app for containers. So you mentioned hybrid cloud, hybrid cloud, multi-cloud. Are the terms this week? We hear them a lot, been up on stage. One way of putting it is thinking about that different places of deploying. But in one way, you're really saying that it doesn't matter where you deploy. There's layers of, and especially OpenShift, can take you to different clouds. Location doesn't matter anymore. Can you drill down on that a little bit? Absolutely, I mean, our whole, we took a bet. I mean, it sounds obvious now. It always does, right? We took a bet on hybrid cloud. I've been talking about it for six or seven years. And what it means is customers are going to have applications that run on bare metal. They're going to have it running as virtual machines, probably on VMware. They're going to maybe run their private clouds, maybe containers, maybe across multiple clouds. End of the day, it's Linux underneath that. What customers don't want is five different operating because every Linux is slightly different. They don't want five different operating environments. They want one. So what we do with RHEL and with OpenShift is we give you that abstraction layer for your application to code once and you can move that app anywhere. I mean, the clouds, the public clouds have brought a tremendous amount of innovation. And I don't want to say this in a derogatory way, but in some sense, they're like a mainframe because they have their stack all the way up to their products are their services. And so you start up a serverless of Lambda. That's never leaving Amazon, never. So it's great in many cases, if that's okay for that app. But there's a lot of cases you might want to run the app here one day and there the next day. So you really need an abstraction layer to ensure that you have that portability. And that's what OpenShift and containers are so important right now. When I hear things like de facto standard and abstraction layers, the bells go off. Ooh, opportunity, because you have now, that's where complexity can be reduced down when you have good abstraction layers. But we've been interviewing folks here and some themes have come up about the sea change that we're facing this cloud scale, new internet infrastructure going on globally. And the two points are TCP, IP moment, during that time that was networking and that disrupted DECnet, SNA and others. And then HTTP, both are different. HTTP was all new capabilities, the web disrupted direct mail and other things analogically. But TCP created internet networking basically. It's just going everything else. Here, what containers was interesting and I want to get your reaction to this is that with containers, I don't have to kill the old to bring in the new. I can do the new and then let the life cycle of those workloads take a natural course. This is a good thing for enterprise, they don't have to rush in, do a rip and replace. They don't have to re-architect higher new people at massive scale. Talk about that dynamic, because that seems to be what's happening. It's exactly what's happening. You know, we did a bunch of demos on stage this week. I think nine of them live. The coolest demo was the one where we showed we actually took a Windows SQL based virtual machine from VMware with tools. We brought that over to a KVM environment, which is a different format for the VM. Brought that to a KVM environment. We then used tools to slice it up into two containers, one being the app itself, the other being SQL, then we deployed it out to OpenShift and we could eventually have deployed out to any public cloud. That's significant for two reasons. First of all, you're now seeing Kubernetes orchestrating VMs right beside containers, so you can kind of see where that's going. So that's interesting for the operators now, because now they bring, whittle down some of that complexity. It's really interesting for the developers, because from a developer perspective, they're going to be asked to bring these traditional virtual machines into containers. In the old world, they have to go to a VM where front end to do that, then they have to come over here to a rev or a rel front end to do it. Now they can just bring their VM with tools over, work on it, split it up into containers and deploy it. It's efficiency at its best. And shift without any effort. Without any effort, really. Talk about the impact of the customers, because this is to me the big money moment because that means an enterprise can actually progress and accelerate their digital transformation or whatever they got going on to a new architecture, a new internet infrastructure. We hear things like network effect, decentralized storage with blockchain, new capabilities that aren't measured by traditional older stacks that we've seen on e-commerce, DNS and other things. So a shift's happening. A shift's happening. It's a cloud scale. Synchronous, how are these, how are these scalable? Whole new way. What does that mean for customers? What it means for customers, there's two things that are important. The shift is happening. You're getting tools and platforms to make that shift more seamless. And I'd love to say it's all redhead engineers that are giving this, but the reason why it's moving so fast is because it's open. So the innovation comes from anywhere. It's way too big of a problem for anyone customer to solve. We're just helping our customers consume it. That's one thing. But I think the other thing that's important is not every application is going to be suited to go to a container-based application. So because it's all on that rel common layer, our customers can still have one operating environment and have compatibility as they do the shift, but still keep their business going over here, maybe forever. These apps may never come off of bare metal, for example. Paul, I wanted to talk a little bit about Red Hat's scope inside IT. I love the connection between the container layer, that is just Linux, and also the standards layer. But now that we're up with OpenShift and with multi-cloud, global, huge scale operation, there's a lot more involved. Cloud-level ops is now, Red Hat is involved with process and culture and you have a lot more that you're involved in helping IT with than just Linux and connection to the back to the machine. So can you talk a little bit about what you're trying to do with the customers? It's a great point. When I started with the company 17 years ago, we weren't talking to CIOs. In fact, the CIOs, we were coming in the back door. The operations people were bringing Linux in the back door and the CIOs didn't even know it was running their environment. But now, as you said, where CIOs are trying to figure out how does public cloud fit into my IT environment? How does a multiple public cloud fit in? How do containers fit in? What do I do with my older applications? They're re-architecting. That's at the CIO level now. They're having to re-architect for the next generation computing. So we've had to build services around that. We have labs, we have innovation labs. We bring our customers in and work with them and help them figure out and help them map out where they're going. For the first time, we actually, I've had many customers tell me this is the, with OpenShift, it's the first time I've got my developers and my ops people in the same room and we've facilitated that discussion. It's because it's going to be one motion. And so that's the interesting part for us. We've really moved up the chain in our customer base because we're almost a consultative sale now to help them get to the next generation. Talk about the enabling aspect of this because again, I reference TCPIP and HTTP, but now if you go forward and say, okay, we're going to have this new environment, it's not just about Red Hat, it's about Linux, it's about the operating system which you guys obviously offer for free and then have services around it and have software. How is Linux with the new capabilities with OpenShift and standards like Kubernetes with containers, how in your opinion is that an enabling opportunity for ecosystem, new startups and enterprises themselves? Because if this happens, it continues to happen. It's going to be new names are going to come out of the woodwork, new startups are going to happen. You see it every day. I mean, you wouldn't do a startup today that wasn't software wise, that wasn't based on Linux and that's why and all the innovation today because all the innovation today is based on Linux. You know, one of the things we released last week at KubeCon is I don't know if you saw it or not, we released a Kubernetes SDK and it came with the CoreOS guys. We put that out into the community. It's really an SDK for ISVs and software vendors to build into the APIs of Kubernetes in an open way so that once they get out into the commercial world, they're ready. That's how significant we all think that Kubernetes is going to be. And we think that's where the services are going to hang from in the infrastructure. But having said that, I think it also tells you that the impact that these open technologies are having on the future. I want to get into the CoreOS in a minute but I want to ask you about the white spaces. So if someone who's at in charge of the troops inside Red Hat and products and technologies, where's the white space opportunities that people can dig in and build out innovation around this major shift that you guys are on this wave? Where's the opportunity for the channel partners, the integrators, global websites, developers, anyone? Where's the key areas? I mean, with our platforms of OpenShift and OpenStack, we have certified entry points via APIs and storage, networking, management. So we've got hybrid management but certainly we don't think we're going to do everything in management by any stretch. We have a set of APIs for management partners to plug in. And by the way, what I tell my management R&D folks, no hidden APIs, same APIs we use, they use. So storage is another area, new storage solutions, networking, certainly AI is one of the areas. One of the things we showcased here was AI permeating through our entire product line. I don't know if you saw the face recognition demo out there, but it was pretty cool. And even if you want to consume that AI through one of the cloud providers, we can pass you straight through from OpenShift to consume it that way as well. On automation, I want to get your thoughts on something that we've talked to two days ago here on theCUBE was automation is great. So let's give you an example. I'm automating a service with Kubernetes and containers and there's a memory leak and it reboots, but automates, I don't know. So you got to have a new level of instrumentation down at the code level. How do you see that playing out? Because now we got to be smarter about what's working and not working because I might not never know, it just reboots intermittently, could be some memory leak, could be something else. But that's, so this is one of the places we're using AI. So we've been, our first stint with AI came out of our support group. So we've been supporting Linux and open source for 25 years, got a massive database of what failures were, what the fixes were. We started using AI in a support group to point our reps at a particular article based on the symptoms that they were hearing from our customers. We realized we had about an 80% hit rate on getting our reps to the right article. So now we've built that into the products and so when we use that AI, like for example, OpenShift.io, which is one of our developer platforms, developers trying to link in a library. We can tell them, you know what, there's a newer version of that library. You know what, that library has a security flaw and it's at this line of code, maybe you want to consider using another one. But it's from our years and years of doing this that we're building that database. I mean, oh, AI is only so good as the data that you've fed it. And so- You have level of granularity down to do it and then also AI. It also is a reason why all our services are now on OpenShift because you're absolutely right. If I've got a raw Jboss service running on raw Amazon, I can't instrument that underneath because Amazon's got that layer closed. If I have OpenShift there, and the infrastructure's OpenShift, even running on Amazon or Azure anywhere, we can now instrument that to look at some of the things we need to look at to recognize an event or whatever. Paul, talk about the journey with CoreOSC. Obviously, we've been super excited by that. We've been following CoreOSC from the beginning. Great technical team, pure open source guys. And in that container part of the evolution in time, everyone's trying to force a business model. And really hard to force a business model is something that too early or might not even be relevant to build a business around. It might be a feature, not a company kind of thing. So you guys put a big price tag on them, sizeable chunk of cash. How did it all play out? You guys just like, hey, wow, we like these guys, they're super technical, meeting of the minds, and then how that has fit in from a product and technology standpoint. A little of all of that. Of course, the benefit of having open source development in your DNA is we knew them all, right? So we knew how good they were because we work, our guys work with them every day, so that was something. When they decided early on like us to go to Kubernetes, they became a big part of the community of Kubernetes. In our model from day one, you can't be an open source provider if you're not strong in the upstream community. Because how can you affect what your customers are asking you to do if you can't affect upstream? They were big in the upstream, big in Kubernetes. And so at that point, that's when we decided, they had done some interesting things that we hadn't got to. They did a lot of the automation. They were doing over-the-air updates of the container platforms, which we hadn't got to yet. They had a really good following in the community. So we decided, you know, we paid a hefty price, but at this stage of the game, we really feel that we took an early bit in Kubernetes, we really feel that that's going to be the future in containers. If there's going to be a place that you pay maybe a little more, this is the place. Well, Paul, I think another example is Ansible a year or two back, right? And that's been, we've remained a huge success. And if I can say it, you haven't messed it up, right? And it's been powerful. That's a complication, by the way. Well, most acquisitions, you know, end in tears. So it seems like Red Hat seems to be good at this kind of an open source acquisition. We get to interview them for two years before we bring them in based on how we work in the community. But, you know, we're bringing in people. I hate to say the word MNA or acquisition. I just hate that word because we're just joining forces here. You know, it just took a big check to do it. Yeah, and you guys got the business model nailed down. This is as good was good for Core at the time for them too. They didn't have to worry about having to figure out a go to market and monetize an upstream presence, which was very valuable. And then trying to shoehorn a business model around it. And which is difficult, companies die doing it. Yeah, I mean, I can't think of many that have been that successful at it. I mean, it's a hard thing to do. I mean, look, we've had a great advantage. You know, we've had RHEL in the market for 16 years and it built a base for us. I'm not going to try to kid you on that. And it's the Linux base that everything's getting built around. And so we just keep those principles we've used for the last 16 years. We stay true to them. We could not do a proprietary piece of software now if our lives depended on it. The DNA. Well, how do you handle the growth? Hiring new people, too. That's a challenge we've been talking to folks about on your team and across Red Hat around hiring people. And you got to maintain that ethos. You have to maintain that DNA. How do you guys do that? Is there like a special three-day hypnotic class? Hey, this is how we do it. I have to tell you, it's a bit easier on the engineering side because it's typically engineers that have been working in the community, et cetera. But our business unit side and other pieces where people have been coming out of big companies and they're used to a hierarchical environment, we really take that into account in the interview process. I'll be frank, not everyone makes it through. I mean, Red Hat is, you know, titles really don't matter. Engineering-led company. Yeah, totally engineering, as all should be, by the way. Pious opinion, but okay. So, great to have you on. Thanks for spending the time. I know you're super busy. Couple questions before we wrap up. What are you most proud of as you look back now? I mean, again, it's almost hindsight's 20-20. It looks obvious, these calls. But I interviewed Diane at OpenStack many years ago. Took a lot of heat for that Kubernetes movement. People weren't, it wasn't obvious to a lot of people at that time, the Kubernetes bet. You guys make good bets. Looking back, what are you most proud of that's most significant or you think people should know? Well, that was a seminal moment in Red Hat history, that decision. What, take us through some of the key milestones in your opinion. The first one, there's probably three or four. The first one was going to Ralph because you have to understand what we did. We were completely retail. When I joined the company, we were $50 million in revenue, losing 200. And so we had a retail product. We stopped it to go to Ralph. Literally stopped the product, bet the company move. Second one was JBoss. We were about 300 million in revenue. We paid 425 million for JBoss. That was a big one. The third one, you might not recognize this one, moving from Zen to KVM. Because Zen was going off down the VC world, trying to figure out how to monetize as a company. Somebody in Israel came up with a better model with KVM. The rest of the industry was on Zen. We said as a single player, we're going this way. That was a big bet that I don't even know we recognize the significance of at the time. And then Kubernetes, as I said, we pivoted on that in 2012 or so. And I've got a lot of R&D money in that and it's paying off. What made you go to Kubernetes? Just curious, because the board success, how software's being done at Google, was it the role of containers? Did you guys have that foresight at that time saying, containers is going to have a critical role. We don't want to screw that up. We can bring this in. Were you looking at it from a stack perspective, or was it more of a future scenario? It was, a lot of it was, it's heritage out of Borg and knowing the talent in Google and engineering. And we talked to, we had many, many discussions. We continually do with those guys. So I think it was mostly a technical decision. And what we said was at that point, putting our weight behind it, we just need to make the community successful. So we figured with us in Google, it was a fairly good bet, not a sure bet, but a good bet and that's what made us go there. It was really a technology decision. Final question as we wrap up, for the folks watching who couldn't make it here in San Francisco for Red Hat Summit 2018, what's the big takeaway? What's the product and technology? What's the North Star for you and your team and what are you guys putting as a priority? What's the focus? I think the takeaway from here is, I think it's pretty solid, couple of things are really solid. It's going to be, the future is going to be open source, period, end of story, especially in the infrastructure and application development world. Third thing is, hybrid cloud is the model. It's the only practical way. Not every application is moving to one public cloud tomorrow. And the third thing is for Red Hat, that's the architecture that we build around every day. We guide what products we build, what M&A we do, everything we do is around that model. And OpenShift obviously is centerpiece of all this. Centerpiece for all that. Thank you for coming on. President of products and technology at Red Hat. I'm John Furrier with John Troyer. Stay with us for more live coverage. Our third day of three days of live coverage here. Out in the open, like open source. We're doing our share, bringing you the content. Be right back with more after this short break.