 from San Francisco, it's theCUBE, covering Red Hat Summit 2018. Brought to you by Red Hat. Hello everyone, welcome back. This is theCUBE's live coverage here in San Francisco of Red Hat Summit 2018. I'm John Furrier, co-host of theCUBE with my co-host analyst this week, John Troyer, who's the co-founder of Tech Reckoning, advisory and community development firm. Our next guest is Red Hash. Bala Krishnan is the general manager of OpenStack for Red Hat. Welcome to theCUBE, good to see you. Great to be here. So OpenStack's very hot, obviously with the trends we've been covering from day one. Been phenomenal to watch that grow and change. But with Kubernetes, you're seeing cloud-native to robust communities. You got application developers and you got under-the-hood infrastructure. So congratulations and what's the impact of that? How is OpenStack impacted by the cloud-native trend and what is Red Hat doing there? The best epitomization of that is OpenShift on OpenStack. If you had caught the keynotes earlier today, there was a demo that we did whereby we were spawning OpenShift on bare metal using OpenStack and then you run OpenShift on pop. That's what we kind of see as the normed implementation for customers looking to get to, I want an open infrastructure on-prem, which is OpenStack, and then eventually want to get to a multi-cloud application platform on top of it. That makes up the hybrid cloud, right? So it's an essential ingredient to the hybrid cloud that customers that are trying to get to. And OpenShift's role in this is what? I'm assuming you were asked about OpenShift. OpenShift will be multi-cloud from an application platform perspective, right? So OpenStack is all about the infrastructure. So as long as you're worrying about infrastructure, deployment, management, lifecycle, that's going to be OpenStack's remit. Once you're thinking about applications themselves, the packaging of it, the delivery of it, and the lifecycle of it, then you're in OpenShift land. So how do you bring both these things together in a way that is easier, simpler, and long-standing? Is the opportunity and the challenge in front of us? So the good news is customers are already taking us there. And there's a lot of production workloads happening on OpenStack, but I got to ask the question that someone might ask who hasn't been paying attention in a year or so with OpenStack. Hey, OpenStack, remember that, what's new with OpenStack? What would you say to that person if they asked you that question about what's new with OpenStack? The answer would be something along the lines of, boring is the new normal, right? We have taken the excitement out of OpenStack, the conversations that are on containers. So OpenStack has now become the open infrastructure that customers can bring in with confidence, right? So that's kind of the boring Linux story, but you know what, that's what we thrive on, right? Our job as Red Hat is to make sure that we take away the complexities involved in open source innovation and make it easy for production deployment, right? So that's what we're doing with OpenStack too. And I'm glad that in five years, we've been able to get here. Definitely, I think along with boring goes clarity, right? Last year, the Cube was at OpenStack Summit, we'll be there again in two weeks with you and I'll enjoy seeing you again. The last year there was a lot of containers versus, there was some confusion, like where people got sorted out in their head, oh, this is the infrastructure layer and then this is the app layer. I think now people have gotten it sorted out in their head, OpenShift on OpenStack, very clear message. So a meeting of the community in two weeks, any comments on the growth of the OpenStack community, the end users that are there, the depth of experience, it seemed like last year was great. Everywhere for OpenStack on the edge, it ended, you know, set top devices and pull top devices all the way to OpenStack in private data centers and for various security or logistical reasons. Where is OpenStack today? Yeah, I think the key phrase would be workload optimization. So OpenStack has now evolved to become optimized for various workloads. So NFV was a workload that people were talking about now. People are in, I mean, customers are in production across the globe, you know, Bead, Verizon, or some of the largest telcos that we have in EMEA and APAC as well. The fact that you can actually transform the network using OpenStack has become real today. Now the conversation is going from core of the data center to the edge, which is radio networks. So the fact that you can have a unified fabric which can transcend from data center all the way to a radio and that can be OpenStack is a, you know, a great testament to the fact that the community has rallied around OpenStack and, you know, delivering on features that customers are demanding. Boring is the new norm, I love that because boring implies reliable, no drama, clean, working. If you had to kind of put a priority in a list of the top things just that are still being worked on, obviously the job's never done with infrastructure, it's always evolving, but DevOps certainly shows that with programmability. What are the key areas still on the table for OpenStack that are key discussion points where there's still innovation to be done and built upon? I think the first one is it's like going from a car to a self-driving car. How can we get that infrastructure to autonomously manage itself? We were talking about network earlier, even in that context, how do you get to a implementation of OpenStack that can self-manage itself? So there's a huge opportunity to make sure that the tooling gets richer to be able to not just deploy and manage but fine tune the infrastructure itself as we go along. So clearly you can call it AI, machine learning, implementation on OpenStack to make sure that the benefit is accruing to the administrator. That's an opportunity area. The second thing is the containers and OpenStack that we touched upon earlier. OpenShift on OpenStack in many ways is going to be the cookie cutter that we're going to see everywhere there's going to be private cloud. If you've got a private cloud, it's got to be an OpenShift on OpenStack. And if it's not, I would like to know why. Right, it becomes a de facto standard. You start to have enablement skills training for folks. As you talk to the IT consumer, right, the IT admins out there, you know, what's the message in terms of upskilling and managing say an OpenStack installation and what is Red Hat doing to help them come along? So those who are comfortable with rel Linux skills are able to graduate easily over to OpenStack as well. So we've been missionarily focused on making sure that we are training the loyal Linux install based customers and with the addition of the fact that now the learning offerings that we have are not product specific but more at the level of the individual can get a subscription for all the products that Red Hat has, you could get learning, access to learning. So that has helped make sure that people are able to graduate or evolve from being able to manage Linux to manage a cloud and then face the brave new world of hybrid cloud that's happening in front of our eyes. But Ash, talk about the customer conversations you're having as the general manager of OpenStack Red Hat. What's the nature of the conversations? Are they talking about high availability, performance, or is it more under the hood about OpenShift and containers or are they range across the board depending upon the use cases? Of course they do range but the high road a bit is that applications is where the focus is, workloads is where the focus is. So the infrastructure in many ways needs to get out of the way to make sure that the applications can be moving from the speed of thought to execution, right? So that's where the customer conversations are going. So which is kind of ties back to the boarding is the new normal as well. So if we can make sure that OpenStack is boarding enough that all the energy is focused on developing applications that are needed for the enterprise, then I think the job is done. Self-driving OpenStack means when applications are just running and that self-healing concepts you were talking about, automation is happening. Exactly, that's the opportunity in front of us. So inch by inch, code by code, we will get there I think. I love the demo this morning which showed that off, right? Bare metal stack sitting there on stage from different vendors, right? Actually OpenStack is the infrastructure layer so it's out there with servers from Dell and HPE and others, right? And then booting up and then the demo with Amadeus showing OpenStack and public clouds with OpenShift all on top, also showed how it fit into this whole multi-cloud stack. Is it challenging to be the layer with the hardware? Are hardware heterogeneous enough at this point that OpenStack can handle it? Are there any issues there working with different OEMs? If you look at the history of Red Hat, that's what we've done, right? So the rail became rail because of the fact that we were able to abstract multi-various innovation that was happening at the x86 level. So being able to bring that for OpenStack is like, that's the right to swipe the employee card if you will, right? So I think the game is going back to what you were earlier talking about. The game is evolving to now that you have the infrastructure which abstracts the compute storage networking, et cetera, how do you make sure that the capacity that you've created is applied to where the need is most, right? For example, if you're a telco and if you're enabling 5G IoT, you want to make sure that the capacity is closest to where the customer pool is, right? So being able to react to customer needs or the customer's customer's needs around where the capacity has to be for infrastructure is the programmability part that we can enable, right? So that's a fascinating place to get into. I know you are technology users yourself, right? So clearly you can relate to the fact that if you can make available just enough technology for the right use case, then I think we have a winner at hand. Yeah, and taking, as you said, taking the complexity out of it also means automating away some of those administrative roles and moving to the operational piece of it, which developers want to just run their code on. It kind of makes things go a little faster. And so, okay, so I get that, but I got to ask the question that's more Red Hat specific if you could weigh in on this, because this is a real legacy question around Red Hat's business model. You guys have been very strong with RHEL. The record speaks for itself in terms of warranty and serviceability you guys give life. I mean, how many years is it now? Like in a zillion years support for RHEL? OpenStack is boring. Is Red Hat bringing that level of support now? How many years? Because if I use it, I'm going to need to have support. What's the Red Hat current model on support in terms of versionings and the things that you guys do with customers? Thank you for bringing that up. What we've been consciously doing is to make sure that we have lifecycle that is meeting two different customers' segments that we're talking about. One is customers who want to be with the latest and the greatest, closer to the trunk. So every six months there is an OpenStack release. They want to be close enough. They want to be consuming it, but it's got to be production ready in their environment. The second set of customers are the ones who are saying, hey, look, the infrastructure part needs to stay there, cement it well, and then every, maybe a couple of years, I'll take a real look at bringing in the new code to light up additional functionality around storage or network, et cetera. So when you look at both the camps, then the need is to have a dual lifecycle. So what we have done is with OpenStack Platform 10, which is two years ago, we have a up to five year lifecycle release. So OpenStack Platform 10 was extensible up to five years. And then every two releases from there, 11 and 12 are for just one year alone. And then we come back to, again, a major release, which is OSP 13, which will be another five years. I know it can be- And they get the full Red Hat support that they're used to. That's right. So the idea is that you're able to either stay at 10, or you could be the one who is going from 10 to 11 to 12 to 13. There are some customers who are saying staying at 10, and then I want to go over to 13. And how do you do that will be industry first, and that's what we have been addressing from an engineering perspective. And that's differentiated too. I think that's a good selling point. Congrats, that's always a great thing about Red Hat. You guys have good support, give the customers confidence. You guys aren't new to the enterprise and these kinds of customers. So Red Hatch, what are you doing here at the show? Red Hat Summit 2018. What's on your agenda? What's some of the hallway conversations you're hearing? Customer briefings. Obviously some of the keynote highlights were pretty impressive. What's going on for you? It's all about open shift on open stack. That's where the current and the future is. And it's not something that you have to wait for. The reality is that when you're thinking about containers, you might be starting very small, but the reality is that you're going to have a reasonably sized farm that needs to power all the innovation that's going to happen in your organization. So given that, you need to have an infrastructure management solution taught through and implemented on day one itself. So that's what open stack does. So when you can roll out open stack, and then on top of it, bring in open shift, then you're not only taking care of today's needs, but also as you scale, and back to the point we were talking about moving the capacity where it's needed, you have an elastic infrastructure that can go where the workload is demanding the most attention. So here's another question that might come up if I want to ask you, and you've probably got this, but I'll just bring it up anyway. I'm a customer of open stack, or someone kicking the tires, learning about deploying open stack. I say, Ritesh, what is all this cloud native stuff? I see Kubernetes out there. What does that mean for me vis-a-vis open stack, and all the efforts going on around Kubernetes, and above in the application pieces of the stack? Right. Let's say if you looked at the rearview mirror five years ago when we looked at cloud native as a construct, the tendency was that, hey, look, I need to be developing net new applications. That's the only scenario where cloud native would be thought of. Now fast forward five years now, what has happened is that cloud native and DevOps culture has become the default. If you're a developer, if you're not sort of in that cloud native and DevOps, then you're working on yesterday's problem in many ways. So if digital transformation is urging organizations to drive towards cloud native applications, then cloud native applications require an infrastructure that's fungible and elastic, and that's how OpenShift on OpenStack again, coming back to the point of that's the future that customers can build on today, and moving forward as well. So summarize, I would say, what I heard you say, and correct me if I'm wrong, OpenShift is a nice bridge layer, or not bridge layer, but a connection point. If you bet on OpenShift, you're going to have best of both worlds. That's a good summary, and you got to be betting on Open, first of all, is the first order bet that you should be making. Once you bet on Open, then the question is, you got to bet on an infrastructure choice that's OpenStack, and you got to bet on an application platform choice that's OpenShift. Once you got both of these, I think then the question is, what are you going to do with your spare time? Count all the cash that you're making from all the savings, but also choice is key. You get all this choice and flexibility is a big upside, I would imagine. Redesh, thanks for coming on and sharing your insight on theCUBE, appreciate it. Thanks for letting us know what's going on, and best of luck, see you in Vancouver. Thank you for having me. Okay, so theCUBE live coverage here in San Francisco for Red Hat Summit 2018. I'm John Furrier with John Troyer. More coverage after this short break.