 from Austin, Texas. It's theCUBE, covering OpenStack Summit 2016, brought to you by the OpenStack Foundation and headline sponsors Red Hat and Cisco. Now here are your hosts, Brian Grace Lee and John Walz. And welcome back inside the Austin Convention Center here as we continue our coverage here on theCUBE of the OpenStack Summit 2016. We are live here on day three. And IBM is a solid contributor, key contributor to the OpenStack community on the platform sponsorship level. And it's a pleasure to welcome Jason McGee from IBM who is an IBM fellow and also the CTO and Vice President of IBM Cloud Platform. And television star too, he might say. Jason, thank you for being with us. We certainly appreciate that. We'll get into the TV spot in a little bit. First, tell us about just your week, your focus here and what you've been able to get done and kind of how you feel about the way this community has really expanded. Almost exploded, you might say, in the past five, six. Yeah, so we love this week. I mean, as you know, IBM's a big contributor too and supporter of OpenStack. It's key to our strategy around cloud. We're building our entire cloud platform on top of OpenStack. So this is an important week for us. We have a big team that are key contributors and technical leads on major projects within OpenStack. This is an important week. Have them come, meet the community, figure out what the next steps are for OpenStack and the technical work we're going to go do. This is also a big client event. So we spent a lot of time with clients and with other partners and kind of talking about the evolution of OpenStack. I think the growth has been tremendous. I mean, obviously over the last few years, an event like this has not only grown in physical size, but really changed in makeup from kind of the early days of the foundation where it was mostly the people participating in the project, getting together to meet and talk about where to go to an event that is now, I think, a great blend between users and contributors and vendors who are really coming together to understand how to bring Cloud forward. So I think it's been a great week so far and some really great conversations. So how's that changed the nature of your work then? Or the nature of maybe the work of the community at large then because there has been this evolution of maybe the structure of the organization and then of the meetings, what's happened here in terms of different releases and different projects and all of a sudden, you see this maturation kind of occurring, right? People are growing up a little bit. So how's that changed things at IBM, you think? Well, I think there's a couple of interesting dimensions there. I mean, one is I see a real maturity happening in OpenStack and I think you see that in the event this week. You know, a lot of stories about clients in production, large-scale deployments, operational focus. I think within the working groups, there's been much more focus on how do you actually operate this platform and how do you build the right kind of operational tools, which is a sign of people really using it, right? So we're beyond the what does it do and into the how do I run this efficiently and operate it? That's important for us in IBM. You know, we're running OpenStack at very large scale, both for our public cloud and for dedicated and on-premise environments. So I like that focus. I think that maturity is good. The other thing which I think is quite apparent this week is the role of OpenStack is starting to get kind of better understood within the broader landscape of cloud. You know, I did a talk yesterday as part of our IBM day, where I talked about kind of the overall cloud stack that we're building, right? From infrastructure through platform services with containers and into domain services like Watson and video. And I think in the industry overall, you'll see a more mature conversation about the role that infrastructure with OpenStack plays and how it relates to Kubernetes and Docker and container technologies. And I think that's good. You know, it feels like we're getting into a better place where we're not all just trying to solve all the same problems. And then we kind of understand how these different communities can play off with each other. Yeah, IBM's obviously been involved with Open Source for a very long time. It's kind of new and cool for some of the other vendors in the industry to go, hey, we now do Open Source. We're contributing. How important do you find when you talk to your customers that they go, you not only understand the technology, but you understand the process. We can lean on you to help us. Because I mean, we're seeing, you know, as much customers using, you know, sort of consuming Open Source, but now they're contributing, which is kind of a crazy phenomenon. But yeah, it's a great point. I mean, in fact, we have those conversations all the time with our clients about, you know, not only what are we doing in Open Source, but how do we help them participate? You know, one of the roles I think IBM has played a lot in recent years is helping different open source communities move towards the proper governance models. You know, we helped create the OpenStack Foundation. We helped create the Cod Foundry Foundation. We're doing some of the same things with CNCF and NOCI and the container space. You know, it's how do we take that experience that we've had with how to do governance and contribution around Open Source and help the communities move forward. For clients, you know, we've done things like, we released this thing called the Bluemix Garage Method, which was a way that we actually documented kind of how we do development internally in IBM, how we transformed ourselves, and we documented that so clients could kind of take advantage of those same techniques and tools and kind of learn how to contribute. So we think that's pretty important role that we can play with our clients for sure. And the Bluemix Garage is great. I mean, there's probably about a half dozen or 10 now. It's a really open place. You go in, you bring a problem. There's a bunch of experts there to help you actually write code. And we were at IBM Interconnect a few, about a month and a half ago, whenever it was. One of the things that I heard over and over again was how much people really liked the mix of, you know, sort of local service, bring the whole Bluemix Cloud local and then having it, you know, in the Cloud and the public Cloud as well. How much do you guys find that people are starting to realize this idea of, there's undifferentiated sort of heavy lifting that people like IBM can help them with and they can focus on applications versus before having to think about, I got to figure all this stuff out. I got to run it. Are you seeing that more and more from your clients going, I got to focus on the application. I got to focus on the business side of things? Yeah, absolutely. Especially if you talk to the right people within those organizations. But I do think, you know, this idea that mixed environments or hybrid state is really the reality that people are going to face. It's not some temporary or transitory state that they need the flexibility to run some things in multi-tenant public and some things on-prem. And that running, you know, running a Cloud, running OpenStack, running Cloud Foundry or containers, like running a platform, it's a fundamentally more complex operation than running a Linux server or an app server. And you really think about it like kind of the blast radius for failure is much bigger. You know, if you do that wrong and you have a thousand apps running on your platform and you screw something up, you take a thousand apps down, right? And so I think, you know, we've run into a lot of clients who kind of started their journey in on-premise Cloud space trying to do it themselves and build it themselves. And they quickly realized that that's an expensive and complicated operation. And so they want to vendor like an IBM to be able to help come in and kind of run that for them. That's why if you look at our local strategy, all of our kind of local Cloud offerings are essentially managed. We use this technology called Relay to let us kind of reach in and connect to and manage that environment. So it's still delivered as a service, you know what I mean? Cloud is a service that you deliver. It's not software. And so even in an on-prem case, you need that kind of capability. Yeah, to the point that you raised that when things do go wrong, right? When something happens and maybe that creates a concern for potential clients, or prospective clients, or existing clients, who are looking at making a similar move and they're like, I don't know what happened here. You have some proof in the pudding. You talk about interoperability and some case studies and some examples. I mean, how do you make me feel better about taking that plunge that maybe our guys back in IT are kind of scratching their chins a little bit and saying, I'm not so sure, Jaycee. Yeah, I mean, some of it I think is just about transparency. I mean, part of the value of the whole community process is that all of the experiences, all of the testing, all of those case studies are things that you can share publicly. And so a lot of it is conversations with clients about, here's what other people have done. Here's the tests we've run. This week we made this request to the other vendors within the Opus Act community to really get serious about doing interop testing, right? And to me, one of the big values of open source and these communities is freedom of action and freedom of movement for clients, that they're not locked into a vendor. They are selecting a technology that's owned by a community and they have the freedom to move things around. And that value doesn't really play out unless there is actual true interoperability that everyone's implementations work the same. And the best way to make that happen is actually prove it. To get together, to do that testing in a publicly and transparently share the results. And I think OpenStack's made a lot of strides recently with projects like RefStack to try to start to define within this incredibly flexible world that is OpenStack where you can plug in all kinds of drivers in different configurations. What's the configuration that we can all rely on and have consistent behavior? I think we have to go beyond that now and kind of do proof, right? Get together and prove it by actually testing and sharing those results. And who are the new clients? I've heard a lot about retailing this week and financing, whatever. Are you seeing expansion into those areas or maybe some other areas, some verticals that you hadn't been into before that is starting to draw some interest with regard to open source and OpenStack? Yeah, I mean, you know, cloud is fundamentally a horizontal technology. So on the whole, it doesn't tend to play into one industry or another. But of course, rates of adoption are different. You know, we're certainly seeing adoption across all industries that we play in in retail and financial services and manufacturing across the board. I have seen a shift, you know, for example, in financial services in the last maybe 12 months, I think there's been a big shift in mentality towards a greater acceptance of public cloud and multi-tenant or shared public cloud environments where maybe a year ago, I'd walk into that client, they'd be like, we're never going to run outside our firewall. Today, they recognize they're going to do that and therefore the value of something like OpenStack where I can have the same infrastructure and the same API inside, by data center and out is a big value for them because they're going to have a mix, right now. So you've seen shifts like that. You mentioned interoperability. You were talking about it from the perspective of, you know, get Mirantis and HPE and IBM to, you know, the vendors to work together. Right. Every customer we talk to goes, yes, I recognize it's a hybrid world. I'd like to be able to move around. And then you step back for a second and you go, you've never had to worry about that before, right? You had your network, your environment. How do you, the challenge to me is, I mean, there's obviously an interoperability within, say, OpenStack, but you've got technologies like Docker or containers saying, hey, we're going to make you portable. You've got technology like Cloud Foundry, we're going to make you portable. How do you help customers understand what that portability interoperability thing means from their perspective, which is, you know, I'd like flexibility. I don't really care about the bits. I want flexibility. Yeah, no, it's a great point that I think if you look, you know, more deeply at the full Cloud Stack, there are different technologies within those layers and some are better. You know, containers I think are a better model for packaging applications in a portable way. But what you quickly find is you take that, that kind of packaged application and below it, it connects to the network. It collects into storage and those things have to be made portable as well. And I think in kind of the compute space, you have virtual machines that you could get through Nova and OpenStack, you have containers, you have a bunch of options. But when it comes to networking and it comes to storage, you know, OpenStack is really the most mature definition of how to expose those capabilities in a way that's flexible across different environments. And so what we've done is, and what my focus recently a lot has been is, how do we actually get these communities to come together? I was at the Linux Collaboration Summit last month in Tahoe and I did a discussion there specifically on this. Like, we have all these great projects. They all overlap on some levels. That's fine, except for when you really start to think about how do I pull this whole thing together and get a consistent experience across the whole stack. So what we've tried to do for clients is define the whole stack for them. You know, here's all the components, bring them together and then be able to be prescriptive about, okay, if you need to build a portable app, use containers, use networking in this way, give them some guidance on how to optimize for that portability. So this week's all about OpenStack. OpenStack's great, but at the end of the day it's sort of low-level plumbing, right? It's like you said, it exposes disks and networks and stuff. On the flip side, the other end spectrum is Watson, right? You were in a, we were talking earlier, you were in a commercial around the masters about Watson. Give us a great example of somebody using Watson to do stuff that you just go, wow, that blew my mind. I mean, there's just so many incredible stories around Watson. I mean, the thing I actually think is most interesting is there's the big use cases, but what's really interesting to me about Watson is how easy it is for an average developer to get access to just incredibly powerful capabilities. I mean, I think about, as I was kind of growing up as an engineer, if I wanted to do video image processing or natural language processing, that was a deep scientific space that would have been very hard for me to get my hands on. And now you can come into a platform like Bluemix and you can get Watson APIs. And in two minutes, you can build an incredible app. I did a presentation back at Interconnect for this thing called Hack Summit. There's a virtual conference online. And I mean, my team built an app that did like live sentiment analysis, tone analysis on the whole Twitter stream for the conference, right? And it took like a couple of days to hack together this fairly complex natural language processing app off of Watson APIs. So I think, to me, the real interesting story about Watson is just we've made it so accessible that the ideas that people come up with are pretty endless. And that part's fun, right? It's down here in the trenches in OpenStack, make sure it works, make sure it's highly available, and then let the creativity go crazy with the developers. That's a cool contrast. I think in OpenStack, you can't lose sight of that. I mean, at the end of the day, all of these things are about providing a platform to run apps, and the apps are the driver, right? And this is how do we support the creation of those applications. Well, tell us about the spot though. I mean, for those at home who perhaps are watching in their office and maybe they weren't watching the masters remember Jordan Spieth's big meltdown on Sunday and not the spot, tell us about the creation of that and your role in that and the fun you had with it. Yeah, IBM does some pretty amazing things on the advertising side, and as a sponsor of the masters we put together a whole series of spots around Watson and cloud, and I just happened to get the opportunity to participate in one of those, and I always think those things take forever, and in this case it was like the masters was done Friday and Wednesday. I mean, New York recording voiceover for this spot, talking about what we're doing around data and cloud, and they somehow hack that together pretty quickly and get it online. So did you ever see yourself? Have you ever thought about that I would appear in a spot promoting? You know, it's funny, I talk to my friends that work about that a lot, like when I was an engineer back in college and taking classes on algorithms and data structures, but I really needed it was like psychology and public speaking and acting. Those would have been much more useful skills for me then. Well, I think you put it all together in a pretty nice way. Jason, thanks for being with us. Congratulations on the spot, but more importantly, congratulations to your work in the open source community. Obviously, IBM's in there full speed ahead. Yeah, thank you. All right, you bet. Jason McGee from IBM. We'll be back with more from the Austin Convention Center with our coverage of OpenStack Summit 2016 right here on theCUBE.