 Live from San Francisco, it's theCUBE. Covering Red Hat Summit 2018. Brought to you by Red Hat. Hello, welcome back everyone. This is theCUBE's exclusive coverage here in San Francisco, California for Red Hat Summit 2018. I'm John Furrier, co-host of theCUBE with my analyst co-host this week, John Troyer, co-founder of the Reckoning Advisory Services. And our next guest is Arvind Krishna, who's the Senior Vice President of Hybrid Cloud at IBM and Director of IBM Research. Welcome back to theCUBE, good to see you. Hey, John and John, great to meet you guys. So you're just, you can't get confused, you've got two Johns here. Great to have you on because you guys are doing some deals with Red Hat, obviously the leader at Open Source. You guys are one of them as well, contributing to Linux. It's well documented in the IBM history books on your role and relationship to Linux. So, you know, check, check. But you guys are doing a lot of work with cloud in a way that, you know, frankly, is very specific to IBM, but also has a large industry impact. Not like the classic clouds. I want to tie the knot here and put that together. So first I got to ask you, take a minute to talk about why you're here with Red Hat. What's the update with IBM with Red Hat? Yeah, great, John. Thanks for giving me the time. I'm going to talk about it in two steps. One, I'm going to talk about a few common tenets between IBM and Red Hat, and then I'll go from there to the specific news. So for the context, we both believe in Linux. I think that's easy to state. We both believe in containers. I think that's the next thing to state. And we'll come back and talk about containers because this is a world, containers are linked to Linux, containers are linked to these technologies called Kubernetes, containers are linked to how you make workloads portable across many different environments, both private and public. Then I go on from there to say, and we both believe in hybrid, hybrid meaning that people want the ability to run their workload wherever they want, be it on a private cloud, be it on a public cloud and do it without having to rewrite everything as you go across. Okay, so let's establish those are the market needs. So then you come back and say, and IBM has a great portfolio of middleware, names like WebSphere and DB2, and I can go on and on. And Red Hat has a great footprint of Linux in the enterprise. So now you say, we got the market need of hybrid, we got these two things, which between them have tens of millions, maybe hundreds of millions of endpoints. How do you make that need get fulfilled by this? And that's what we just announced here. So we announced that IBM middleware will run containerized on Red Hat containers on Red Hat Enterprise Linux. In addition, we said IBM Cloud Private, which is the ability to bring all of the IBM middleware in a sort of a cloud friendly form, right? You click and you install it, keeps itself up, it doesn't go down, it's elastic in a set of technologies we call IBM Cloud Private, running in turn on Red Hat OpenShift container service on Red Hat Linux. So now I'm for the first time, if you say I want private, I want public, I want to go here, I want to go there, you have a complete certified stack that is complete. I think I can say we are unique in the industry in giving you this. And this is where the fruit comes off the tree for you guys, because we've been following you guys for years, you know, where's the cloud strategy? And first of all, it's not like, you don't have a cloud strategy, you have cloud products, right? So you have to deliver the goods. You got just a replay. The market need we all know is hybrid cloud, multi-cloud, choice, et cetera, et cetera. You take Red Hat's footprint, your capabilities, your combined install base is foundational. So, and nothing needs to change. There's no lift and shift, there's no rip and replace. You can, it's out there, it's foundational. Now on top of it is where the action is. That's what we're, that's what we're kind of getting at, right? That's correct. So we can go into somebody, they're running a, let's say a massive online banking application or they're running a reservation system. It's using technologies from masters, using Linux underneath. And today it's all a bunch of peace spots. You have a huge complex staff. It's all hardwired and rigidly nailed down to the floor in a few places. And now you can say, hey, I'll take the application and don't have to rewrite the application. I can containerize it. I can put it here. And that same app now begins to work, but in a way that's a lot more fluid and elastic. Oh, by the way, I want to do a bit more work. I want to expose a bit of it up as microservices. I want to insert some AI. You can go do that. You want to fully make it microservices-enabled to be able to make it into little components and digestible, you can do that. So you can take it in sort of bite-sized chunks and go from one to the other at the pace that you want. And that's game-changing. Yeah, that's what I really like about this announcement. I mean, it really brings the best of breed together, right? You know, there's a lot of talk about containers and legacy and we've been talking about what goes where and do you have to break everything up, like you were just saying. But the announcement today, you know, WebSphere, the battle-tested, huge enterprise-scale component, DB2, those things containerized and also in a framework like with IBM, either with IBM microservices and application development things or others, right? That's a huge endorsement for OpenShift as a platform. Absolutely, it is. And look, we would be remiss if we didn't talk a little bit. I mean, we use the word containers and containers a lot. Yes, you're right. Containers is a really, really important technology. But what containers enable is much more than prior attempts such as VMs and all have done. Containers really allow you to say, hey, I solved the security problem. I solved the patching problem, the restart problem. All those problems that lie around the operations of a typical enterprise can get solved with containers. VMs solved a lot about isolating the infrastructure, but it didn't solve, as John was saying, the top half of the stack. And that's, I think, the huge power here. Yeah, I want to just double-click on that because I think the containers thing is interesting to me because, you know, first of all, being in the media and loving what we do, we're kind of a new kind of media company, but traditional media has been throwing IBM under the bus and saying, whoa, you know, old guard and all these things, but here's the thing. You don't have to change anything. You could, containers, you can essentially wrap it up and then bring a microservices architecture into it so you can actually leverage at cloud scale. So what interests me is that you can move instantly, value proposition-wise, pre-existing market, cloudify it, if you will, with operational capabilities. And this is where I like the cloud private. So I want to kind of go there for a second. If I have a need to take what I have at IBM where there's web sphere, no, I got developers, I got install base, I don't have to put a migration plan. I containerize it. Thank you very much. I do some cloud-native stuff, but I want to make it private. My use case is very specific. Maybe it's confidential. Maybe it's like a government region, whatever. I can create a cloud operations. Is that right? I can cloudify it and run it. Absolutely correct. So when we look at cloud private to go down that path, we said cloud private allows you to run on your private infrastructure. But I want all these ability to just describe John. I want to be able to do microservices. I want to be able to scale up and down. I want to be able to say operations happen automatically. But it gives you all that, but in the private, without having to go all the way to the public. So if you care a lot about, you're in a regulated industry because you went down government or confidential data, all you say, this data is so sensitive, I don't really, I'm not going to take the risk of it being anywhere else. It absolutely gives you that ability to go do that. And that is what we brought cloud private to the market for. And then you combine it with OpenShift and now you get the powers of both together. So you guys essentially have brought to the table the years of effort with BlueMix, all that good stuff going on. You can bring it, you can actually run this in any industry vertical, pretty much, right? Absolutely. So if you look at what the past has been for the entire industry, it has been a lot about constructing a public cloud. Not just us, but us and our competition. And a public cloud has certain capabilities and it has a certain disaster city. It has a global footprint, but it doesn't have a footprint that's in every zip code or in every town or in every city. That's not going to happen to the public cloud. So we say it's a hybrid world, meaning that you're going to run some workloads on a public cloud. I'd like to run some workloads on a private and I'd like to have the ability that I don't have to pre-decide which is where. And that is what the containers, the microservices, the OpenShift, that combination all gives you to say you don't need to pre-decide. You rewrite the workload onto this and then you can decide where it runs. While I was having this conversation with some folks at a recent Amazon web services conference, I said, well, if you go to cloud operations, then the on-premise is essentially the edge. It's not necessarily, then the definition of on-premise really doesn't even exist. So if you have cloud operations in a way, what is the data center then? It's just a connected tissue. That's right. It's the infrastructure which you set up and then at that point, the software manages the data center as opposed to anything else. And that's kind of been the goal that we've all been wanting. So it sounds like this is visibility into IBM's essentially execution plan from day one and seeing and connecting the dots, having the ability to take either pre-existing resources, foundational things like Red Hat or whatnot in the enterprise, not throwing it away, building on top of it and having a new operating model with software, with elastic scale, horizontally scalable, synchronous, all those good things, enabling micro-search with Kubernetes and containers. Now for the first time, I could roll out new software development life cycles in a cloud-needed environment without foregoing legacy infrastructure and investment. Absolutely. And one more element, and if you want to insert some public cloud services into the environment, be it in private or in public, you can go do that. For example, you want to insert a couple of AI services into the middle of your application, you can go do that. So the environment allows you to do what you described and these additions. I want to talk about people for a second. The titles that we haven't mentioned, CIO, business unit leaders, how are they looking at digital transformation and business transformation in your client base as you go out and talk to them? So let's take a hypothetical bank. And every bank today is looking about at simple questions. How do I improve my customer experience? And everyone when they say customer experience, they really do mean digital customer experience to make it very tangible. And what they mean by that is, how do I get my end customer engaged with me through an app? The app's probably on a device like this. Some smartphone, we won't say what it is. And so how do you do that? And so they say, well, well, you obviously want to check your balance. You obviously want to maybe look at your credit card. You want to do all those things, the same things we do today. So that application exists. There is not much point in rewriting it. You might do the UI up, but it's an app that exists. Then you say, but I also want to give you information that's useful to you in the context of what you're doing. I want to say you can get a 10-second loan, not a 30-day loan, but a 10-second loan. I want to make an offer to you in the middle of you browsing credit cards. All those are new customers of this thing. So where do you construct those apps? How do you mix and match it? How do you use all the capabilities along with the data you've got to go do that? And what we are trying to now say, here is a platform that you can go all that, do all that on. Right in that complete lifecycle, you mentioned the development lifecycle, but I got to add to the data lifecycle, as well as here is the versioning, here are my AI models, all those things built in into one platform. And scales are huge, the new competitive advantage, you guys are enabling that. So I got to ask you on the question on multi-cloud. Obviously, as people start building out the cloud on-prem and with public cloud and the things you're laying out, I can see that going on for a while. A lot of work being done there. We've seen that Wikibon had a true private cloud before I thought was truly telling a lot of growth there still not going away. Public cloud certainly has grown, the numbers are clear. However, the word multi-cloud is being kicked around. I think it's more of a future state obviously, but people have multiple clouds. We'll have relationships with multiple clouds. No one's going to have one cloud. It's not a winner-take-all game. Winner-take most, but you're going to have multiple clouds. What does multi-cloud mean to you guys in your architecture? Because is that moving workloads in real time based upon spot pricing indexes? Or is that just co-locating on clouds and saying I got this app on that cloud, that app on that cloud, control plane, these are architectural questions. It's a data. What the hell is multi-cloud? So there's a today and then there is a tomorrow and then there is a long future state, right? So let's take today, let's take IBM. We run Salesforce, we run ServiceNow, we run Workday, we run SuccessFactors. Well, all of these are different clouds. We run our own public cloud, we run our own private cloud and we have traditional data center. And we might have some of the other clouds also through apps that we bought that we don't even know. Okay, so that's just us. I think every one of our clients is like this. So multi-cloud is here today. I begin with that first simple statement. And I need to connect the data, I need to connect when things go where. The next step, I think people, nobody's going to have only one even public cloud. I think even amongst the big public clouds, most people are going to have two, if not more. That's today and tomorrow. Your channel partners have clouds, by the way. Global SIs all have clouds. Yeah, all have clouds. The Cube is a cloud for Crane Out Loud. So then you go into the aspirational state and that may be the one you said where people do spot pricing. But even if I stay back from spot pricing and completely dynamic and I'm worrying about network and I'm worrying about video reach, I just back up on two. But I may decide I have this app, I run it on private. Well, but I don't have all the infrastructures. I want to burst it today. And where do I burst it to? I got to decide which public and how do I go there? And that's the problem of today. And we're doing that. And that is why I think multi-cloud is here now. Not some point in the future. The problem statement there is latency, managing service level agreements between clouds and so on and so forth. Managing access control on governance. Where does my data go? Because there may be regulatory reasons to decide where the data can flow and all those things. That's a great point about the cloud. I never thought about that way. It's a good illustration. I would also say that I see the same argument in the database world. Not everyone has DB2. Not everyone has Oracle. Not everyone has, databases are everywhere. You have databases, part of IoT devices now. No one makes a decision on the database. Similar with cloud. You're seeing a similar dynamic. It's the glue layer that interests me. How do you bring them together? So holistically looking at the 20 mile stair in the future. What is the integration strategy long term? If you look at a distributed system or an operating system, there has to be an architectural guiding principle for integration, your thoughts. It doesn't mean a world that's 30 years in the making. So we can say networking. Everybody had their own networking standards in the, let's say the 80s, so it probably goes back to the 70s, right? You had SNA, you had TCPIP, you had NetBIOS. Decknet. Decknet, you can go on and on. And in the end, is TCPIP that won out as the glue? Others, by the way, survived, but in pockets and then TCPIP was the glue. Then you can fast forward 15 years beyond that and HTTP became the glue. We call that the internet. Then you can fast forward and you can say, now how do I make applications portable? And I would turn around and tell you that containers on Linux with Kubernetes as orchestration is that glue layer. Now in order to make it so, just like in TCPIP, it wasn't enough to say TCPIP, you needed routing tables, you needed DNS, you needed name repositories, you needed all those things. Similarly, you need all those here, I call those catalogs and automation. So that's the glue layer that makes all of this work. This is important. I love this conversation because I've been ranting on this on the queue for years. You nailed it. A new stack has developed in DNS. This is old internet infrastructure. Cloud infrastructure at the global scale is seeing things like network effect. Okay, we see blockchain and token economics, databases, multiple databases on structured data. So a new plethora of new things are happening that are building on top of, say, HTTP, and this is the new opportunity. This is the new platform which is emerging and it's going to enable businesses to operate. You said at scale to be very digital, to be very nimble, application life cycles are not always going to be months, they're going to come down to days and this is what gets enabled. So I want you to give your opinion, personal or IBM or whatever perspective, because I think you nailed the glue layer on Kubernetes Docker and this new glue layer that, and you made reference system, things like HTTP and TCP, which changed the industry landscape. Wealth creation, new brands emerged, companies we've never heard of emerged out of this and we're all using them today. We expect a new set of brands are going to emerge, new technology are going to emerge. In your expert opinion, how gigantic is this swarm of new innovation going to be? Because you've seen many ways before. In your view, in your mind's eye, what are you expecting? What do you share your insight into? How big of a shift and wave is this going to be and add some color to that? I think that if I take a, I'll take a shorter and then a longer term view. In the short term, I think that we said that this is on the order of $100 billion. That's not just our estimate, I think even Gartner estimated about the same number. That'll be the amount of opportunity for new technologies in what we've been describing. And that is, I think, short term. If I go longer term, I think as much as a half, but at least a fourth of the complete IT market is going to shift onto these technologies. So then the winners are those that make the shift and then by conclusion, the losers are those who don't make the shift fast enough. Of the market moves, that's huge. It's interesting, we used to look at certain segments going back years. This company's re-platformizing, re-platforming their lift and shift and all this stuff. What you're talking about here is so game changing because the industry's re-platforming. It's not a company, it's an industry. That's right, and I think the internet era of 1995 to pick that point is perhaps the easiest analogy to what is happening. Not the emergence of cloud, not the emergence of all that. I think that was small steps. What we're talking about now is back to the 1995 statement. Every vertical is upgrading their stack across the board from e-commerce to whatever. That's right. It's completely modernizing. Correct. Around cloud. What we call digital transformation in a sense, yes. I'm not a big fan of the word, but I understand what you mean. Great insight, Armin. Thanks for coming on theCUBE and sharing the press. We didn't even get to some of the other good stuff, but IBM and Red Hat doing some great stuff. Obviously, foundational. I mean, Red Hat, Tier One, first class citizen in every single enterprise and software environment. Now, open source runs the world. You guys are no stranger to Linux being the first billion dollar investment going back. So you guys have a heritage there. So congratulations on the relationship. I think 18 years ago, if I remember, 1999. Yeah, and I love the strategy. Hybrid cloud here at IBM and Red Hat. This is theCUBE bringing all the action here in San Francisco. I'm John Furrier, John Troyer. More live coverage. Stay with us here in theCUBE. We'll be right back.