 Live from Santa Clara in the heart of Silicon Valley, it's theCUBE, covering Cloud Foundry Summit 2017. Brought to you by the Cloud Foundry Foundation and Pivotal. Welcome back, I'm Stu Miniman with my co-host John Troyer. Happy to welcome back to the program a multi-time guest. Stephen O'Grady who is principal analyst and co-founder of Red Monk, very good to see you. That's my pleasure. Yeah, we said those of us in the Boston area, we have to come cross country to be able to talk sometimes. Quite the flight. So, you know, you at Red Monk obviously heavily involved with the developer community. You've been at this show a number of times. What's the vibe? What are some of the kind of big themes that you've been seeing at the show and any big changes from previous years? I think the biggest change or probably the biggest theme that I see this year is that it's more of a developer focus than we've seen in the past. So Red Monk, you know, we use the term developer but we use that term loosely. It can refer to a variety of practitioners, right? In some cases it might be ops staff and other cases might be, you know, people administering a database. In other cases it might be, you know, sort of people actually writing code. And certainly I think from a theme perspective and from an attendance perspective, you know, a lot of people that I've talked to here are, you know, a lot of them still have their hand in ops, certainly. But, you know, we're seeing a lot of development attention here. One of the things we've looked at is, you know, cloud foundry in general and Pivotal specifically, selling a lot of big enterprises, you know. This is not, you know, I'm sure there are cases but most of the people we've been talking to, I mean, you know, big insurance companies, Express Scripts, you know, things like that. This is not, oh, you know, small group of developers that are just tinkering and doing some startup but big companies with lots of apps. How does that fit into your view of, you know, where things happen and how transformation happens? Honestly, I think, you know, there's a couple of things that are going on, you know, in particular with respect to developers, you know, basically the thing that we hear from the enterprises all the time is that we want to give developers the opportunity to be productive and focus on the things that they care about, right? So it's not, you know, in other words, this isn't like, for example, a new programming language that enters the enterprise because somebody says, oh my God, I love this new programming language and that's what I'm going to do. But, you know, a lot of the driver here, you know, from, at least from the enterprises that we talk to, right, is about making developers' lives easier. And simpler. And I think beyond that, you know, really when we start to look at, you know, the adoption and sort of where it's going, where it sits in the marketplace, you know, it's, you know, I think very, I think it's very consistent, you know, with what we would expect, you know, which are, you know, for example, environments that are, you know, tend to remove some of the operational concerns. So, all right, you know what? What I really want to do is, you know, build an application. I don't want to necessarily worry about everything that goes into that, you know, everything that has to sort of go on to make that happen and be deployed and be kept healthy. And, you know, just as importantly, I want this to be done in a way that allows me to work the way that I want. So, not single language, not tied to a particular platform that allows me to kind of go wherever I want. So, yeah, I think in terms of the adoption that we see, you're right, you know, there's a lot of enterprise traction, big, you know, sort of traditional businesses. But it's very much, like I said, it's consistent with certainly what we expect for the marketplace. A lot of times, the analyst coverage or the press coverage gets framed in terms of the horse race. Who's winning, who's losing, what's hot, hype cycle. And I don't know if that's always particularly productive and we're probably not in a winner-takes-all sort of even scenario. Is there a better way of looking at this or how should we be looking at platforms and all that? Yeah, I think it's true. Like, everybody wants to turn it into some zero or some equation and more importantly, they want to reduce it to usually A versus B, right? They want a simple binary comparison. I think the way that we look at it is that, you know, when I, like years and years and years ago, when, you know, I was a systems integrator, your choices were pretty minimal, right? There's only one real, like, I shouldn't say one approach. You basically were going to choose between Java, typically, or, you know, the Microsoft stack. Assuming that you were not a Microsoft shop, you go to the Java route, there was basically one way you built applications. It was, you know, sort of three tier architectures, you know, app server, you know, relational database, blah, blah, blah. And that's, that doesn't, like, anything like the world that we see today. You know, what we see today is much more of a, I guess the best way to put it would be sort of different tools for different jobs. And, you know, it is, at virtually every layer, you know, from database, you know, sort of on up to platform to, you know, tooling, you know, anything that goes into it. You know, frameworks, languages, take your pick. You're going to have more choices, you're going to have more diversity, you know, certainly than we did 10, 15 years ago. And as a result, you know, it becomes, you know, again, like I said, just sort of a different tool for different jobs, as opposed to, this is a zero sum equation and, you know, sort of one of these players is going to take all the market because, look, they take different approaches. You know, there's a reason that you have more than one choice. And that's, you know, typically because, you know, like, I might prioritize this, you might prioritize that, in which case, okay, I'm going to pick a different tool then. Right, but people, yeah, people in the ecosystem, using that tool, actually very productive, very happy for this whole generation. Yeah, very much. I mean, but so speaking of different ecosystems, big news in, at this conference, talking about Project Kubo, bringing Kubernetes on top of Bosch, with some of the same tooling, but, you know, being able to install, manage, upgrade. How should people be thinking about, now, the introduction of Kubernetes into the Cloud Foundry platform? Honestly, I think it's, I think, again, it's very consistent with what we just talked about, different tools for different jobs, right? You know, so in other words, you do have people who want to operate at an elemental level with containers, and that is sort of their base building block, that's really the only thing that they care about, in which case, that's great. You know, Kubernetes is a very solid, you know, sort of option for you, and having the ability to leverage, you know, some of the same underlying componentry, as a project like Cloud Foundry, which tends to be more app-centric, right? I'm not thinking that's only in containers, I have an app, and I want to let Cloud Foundry, sort of, do all the heavy lifting and worrying about, you know, hey, what are these containers? So, you know, to me, it's, I think it's a welcome development. So rather than forcing customers into a, you know, sort of particular choice, and trying to remove choices from them, it's basically saying, look, if this is, you know, sort of your prioritization, this is how you look at the world, great, you know, go Kubo, you know, we still have, you know, obviously, sort of a different approach and a different tool for a different job in Cloud Foundry, you know, so I think it's, you know, from my perspective, again, I'm not a member of any one of these communities, but from the perspective of an analyst, it's the kind of thing where, you know, you want to support that kind of choice for customers. How are customers dealing with the paradox of choice? You know, one of the things that you look at is, isn't it much easier if I just choose a platform, it's totally opinionated, it'll help, you know, simplify my environment? I mean, Cloud Foundry in some ways is, you know, opinionated, but, you know, we keep talking about, you know, there's choices, there's options, the SAP keynote yesterday put up this really complicated stack drawing and when he said I'm going to walk through it, the entire audience cracked up a lot, so we know, this isn't easy stuff, so how do customers sort through all that, make the right decisions, you know, stay flexible and yet agile? It's hard, honestly, and I don't want to give my keynote tomorrow away, but that'll be a big, you know, sort of topic of it, but basically, the gist of it is essentially that choice to me is kind of like a parabola, right? So in other words, if you have no choice, that's a bad thing. The more choice, you know, you have, say, you have one choice, two choices, three choices, you know, that goes up, you know, from a utility and a value standpoint, but you know, it begins to go down, right? You begin to see diminishing and in fact, negative returns because if all of a sudden I'm going from, hey, I have to pick between two or three options to two or three projects, to picking between 10 and 20, you know, that quickly, it just doesn't scale. And, you know, particularly these days, as we were talking about earlier, it's not just that I have to do that in one category, I have to do that in every category, right? So take databases that used to be, this is one of the examples I'll talk about tomorrow, used to be there were no other databases. I mean, there were, but you know, 99% of the time is going to be relational database. So the choice of what style database didn't exist. And even within that, you had, you know, two or three commercial options that took the bulk of the market share. So that's not a particularly complicated choice. And look at the market today, quite a bit more complicated. So yeah, in a perfect world, you know, certainly from the CIO's perspective, do they want to standardize on, you know, single platforms, single approach, single language, single database, et cetera? Sure they do. But I think what we're seeing from a marketplace standpoint is essentially a, you know, companies I think are striving to get to a happy medium where it's okay, you know, we want, you know, certainly more choice than just the single platform, one size fits all. This is the solution to all known problems. But we also don't want to go down the, you know, sort of paradox of choice, which is, you know, hey, I'm picking between half a dozen projects in eight different categories, and now I have to figure out how to wire all that stuff together. So in other words, you know, going back to what we were just talking about, you know, if you go to a customer and say, like, you know, if you're app-centric, great, you're Cloud Foundry, if you are, you know, container-centric, you know, there's Kubernetes, and by the way, they're running the same infrastructure, that's not bad, right? You're basically trying to come at it, you know, from a, you're trying to provide choice, but choice in a sort of manageable, sort of bite-sized fashion, you know, as opposed to, you know, hey, here are 50 different projects that you don't have to sort through. What about the discussion of really the multi-Cloud world? How do customers figure out whether they just use, you know, simple infrastructure as a service or go deeper into, you know, services that are, I guess you would say a little bit more proprietary, therefore it's a little bit more sticky, but, you know, makes it difficult to change, seems to be kind of a fundamental challenge for something like Cloud Foundry to play into that type of environment. Yeah, it's a good question, you know, and I think the first step, honestly, for many enterprises is to acknowledge the fact that they're running in Clouds, you know, because we still go out and talk to, you know, a variety of different businesses, right? We say, you know, hey, you know, what Cloud, you know, sort of platforms are using now, and they'll say, oh, we're not using any of that. I'm like, yeah, you are. Yeah, yeah, yeah, wait, when I go speak at conferences, it's like, you know, there's two types of people out there, people that know that they're using Amazon and the people that haven't figured out that somebody was using Amazon. That's exactly right, yeah. So, you know, I think, like I said, I think step one for most of them is essentially to get a handle on, okay, what are we actually using? Step two is, you know, trying to rationalize and sort of have a strategy in terms of, okay, what are we using, how are we using it, and how do we diversify our bets? Because, you know, for a lot of very good reasons, you know, enterprises are reluctant to put all of their eggs in one basket, you know, from a platform standpoint. Which means, okay, you're probably going to use more than one Cloud. Even if the bulk of your work is on Amazon or any other provider, you probably want somebody else to keep them on us, and to essentially hedge your bets, you know, about future price increases or, you know, decreased levels of service, whatever it might be. To enable that, however, you do need essentially layers of essentially insulation from proprietary services, right? And that traditionally, you know, people forget this now, but we go back to, you know, the Java middleware market, that was the purpose of Java middleware, right? All of a sudden I had a layer where, you know, I don't care what the operating system is and I don't care what the hardware is underneath it. If I'm running on BEA WebLogic or IBM WebSphere, basically I can, you know, if I want to, you know, switch from, you know, a Windows box to Unix in the future, you know, or Linux later, sure, I can do that. So basically what we look at with things like Cloud Foundry are, you know, a similar level of sort of insulation and isolation from, you know, being sort of wetted, you know, two proprietary features unique to any one of these platforms. Now, that being said, it's still incumbent on you to make smart choices in terms of the things that are around those platforms. So as an example, if I go with the Cloud Foundry and one of these platforms, and then I sort of layer on top of that or alongside of that, a whole bunch of proprietary features, you know, might be something like Kinesis or something on Amazon or, you know, BigQuery, you know, from Google or something you can only get from those providers, then I've basically undone a lot of the sort of platform-independent work that I've tried to do with Cloud Foundry. So, yeah, it's a multi-step process. It's not a simple process, but, you know, frankly the smart enterprises now are, you know, certainly embracing Cloud, you know, because it offers so many advantages in terms of, you know, being dynamic and speed of provisioning, but you have to think about sort of long term, how do I, you know, sort of make myself at least somewhat independent, you know, from any given provider? So you brought up some of these tensions that are going on, but we, big announcement also this week, Microsoft joining the Cloud Foundry Foundation. We've got Google Cloud here, you know, Amazon, a lot of people also running on Amazon. Yeah, I mean, you've been following the foundation in the ecosystem for a while. I mean, what do you see in terms of the ecosystem? A lot of the big players are not here this year and then kind of have ramped down their Cloud Foundry initiatives. How do you see the health of the ecosystem and where do you see it going? Those are good questions. I think, you know, the joining of Microsoft, I think is a big one, you know, because Microsoft, you know, one of the conversations, for example, you have around the Azure platform, is that Microsoft is in a lot of accounts. And, you know, that has been a huge advantage for that platform really since the day it was born. So I think that announcement was a good one, you know, for the overall health of the ecosystem. But, you know, I think, you know, sort of taking a step back in terms of, you know, looking at Cloud Foundry and where it sits, honestly, I think that, you know, really the most accurate read from my standpoint, you know, it goes back to, you know, one of the first and second questions that we talked about, which is, you just have different tools for different jobs, right? At the time that Cloud Foundry had launched, right? You know, one of the things that was pretty much undeniably true was that if you're looking for essentially Paz platforms, you know, what we called Paz at the time, you either had the choice of, you know, essentially fixed environments, you know, like a GAE, a Google App Engine, you know, Heroku, force.com, you know, players like that, where they would take care of a lot of the heavy lifting for you, but you didn't have any control. You don't really make any of the choices. So Cloud Foundry comes out and basically is, you know, for a while, it's the only game in town if you were thinking about sort of how you build a modern platform. Well, now we just have more options, right? In other words, you know, going back to where we're talking about with Kubernetes, you know, if I have a, you know, sort of container-centric worldview, well, you know, I might want to go a different option or a different direction, you know, than Cloud Foundry, but, you know, honestly, I kind of liken it to every year for a number of years, I would go out to a conference called FOSDA, it's in Brussels, it's a huge open source conference, and I would talk to the Java developers there, and one of the things I would tell them was that you are never going to be as popular on a share basis as you once were, and they'd get very upset, right? You know, a lot of people would be like, ah, I can't believe it, no, we're still everywhere, and there were all these projects and so on, and you'd say, you absolutely are. But if we, again, if we, you know, contrast where they were 10 years ago, where basically everything, as we were talking about, was either .NET or Java, to a world today where .NET and Java are still big properties and use in lots and lots of places, but, you know, people want to use JavaScript, they might want to use Python, they might want to use Go, they might want to use all sorts of alternative languages, so what does it mean? It means that in a marketplace we have multiple choice, you know, each one of those players is probably going to, you know, sort of have less attention and, you know, less, I don't know, I guess overall visibility, if you will, than if you're the only player in that market, so I honestly don't read any more into it than that. You know, I think, you know, frankly, as we were talking about before, some measure of choice is good, and, you know, Frank, I think if you talk to a lot of people in the Cloud Foundry community, they would probably tell you that, you know, hey look, it's good to have some choice, and, you know, conversely, like the people who are, you know, coming at it from a container standpoint, they probably don't want the only choice to be, you know, containers either, you know, if you want options. What, want to get your update on open source business models, you kind of look today, Pivotal's really an open core, look at a lot of the cloud providers today or really delivering open source as a service, how open they are is debatable. One of the vendors uses a term called open cloud, which I've, you know, taken, you know, some, not saying I didn't necessarily agree with it, so you look at, like, your book, The Software Paradox, Software is Even in the World, Open Source is Eating Money. Do you think that's changed much? How do you, how are you seeing the monetization of open source and privilege? I think from my standpoint, and a lot of people will not agree with me when I say this, the monetization, the future of monetization for open source, in my opinion, is largely going to be in the cloud, you know, as a service, right? And it doesn't matter to, like, take a pic, whether it's a platform like a Cloud Foundry, whether it's a database like a Postgres for MySQL, like whatever it might be, that is the simplest, not to execute because it's challenging, obviously to run a service, much more challenging than it is to just write and distribute software. But in terms of the business model itself, it's the cleanest because, you know, even the best open source companies, commercial open source companies in the world, convert a very, very small percentage of their users to actual paying customers. And that's been a problem that everybody has tried to solve using various sort of, you know, wonky, at times awkward licensing mechanisms, you know, it used to be dual licensing with MySQL, the most common today is open core. So, you know, the question always has been, how do you get people to pay for something that's free? And you know, the answer is it's hard, you know? You can work really hard at it, you can build a business and commercial open source organizations are going to have a, you know, in my opinion, a long future. But the real growth and the real opportunity is going to come from the people who don't just try to sell the software, but sell the software as a service. So instead of going to a business and saying, hey, do you want to buy this thing for me for free and then run it yourself, you go to them and say, do you want to buy this thing for me and by the way, I could run it better than you can because I wrote it. So, like I said, a lot of people in a lot of different areas, both analysts and, you know, people actually write software, probably wouldn't agree with that. But again, you know, doing what we do and having the conversations we have and looking at the numbers we do, that just seems the, you know, clearest long-term projection from a model standpoint. Want to give you the final word, other than of course, watching your keynote for people that didn't make it to the show any takeaways for them, you know, update to the communities, things that you chair. Honestly, I think the biggest one for me is I think what we talked about at the top, right? Which is Cloud Foundry community in particular, you know, has been a very sort of ops-centric historically, you know, focused largely on people who operate these platforms and run these platforms sort of at scale. And I think that both the conference here, the foundation itself, the, you know, a lot of people sort of in the Cloud Foundry orbit are thinking a lot more about, okay, you know, hey, the ops side here is really great and that story is important to tell. But we also need to have a good story and a good narrative and a good message for the people who are actually using that platform to build software. So I think that's something that I'd, you know, whether you're here or not, you know, whether you go watch the talks is something I'd pay a lot of attention to. Kiva Grady, always a pleasure to catch up with you. Thanks so much for joining us. For John and myself, we'll be back with more coverage here from the Cloud Foundry Summit here in Santa Clara. Thanks for watching theCUBE.