 OK, I think we're going to go ahead and get started. Is everyone ready? OK, first I'd like to say good afternoon and thank you for attending. Also, thank you to our panelists for joining our fourth OpenStack Summit Media and Analyst Q&A. For those of you who have not attended one of these in the past, I don't know how many of you have been to a media analyst panel at OpenStack before. We've got an esteemed group here who are prolific. They write about the subject probably better than anybody. And several of our panelists have even been here from the beginning. So should be exciting and interesting panel. I also want to leave up an opportunity towards the end for you to ask questions. So keep them and let us know. And we'll leave at least five minutes at the end to get that done. So first, let me introduce myself. My name is Kelly Andreri. I work with Piston Cloud Computing as their director of marketing. And again, rather than doing it in justice, I'm going to have the panelists go ahead and take a minute to introduce themselves to you. And then we'll go ahead and get started with questions. So would you like to know? OK, hi. My name is Rene Bust. I'm a senior analyst and cloud practice lead from Chris Research from Germany. And for Chris, OpenStack is a very hot topic, especially in Germany. We did the very first OpenStack study based on this market. And if you're interested in, of course, you can reach me out. My name is Frédéric Lardinois, in despite the fact that it sounds really French, I'm German. But I write for TechCrunch. I cover a lot of the cloud, cloud infrastructure companies for us and occasionally enterprise. I'm kind of the newbie here. I only started covering OpenStack about half a year ago. So once we lost one of our enterprise writers and I had to jump in. So that's all I got here. Hi, good afternoon. My name is Al Sadowski. I'm a research director with four or five one research. We're one of the top four IT research and advisory firms. I'm based in New York. And this is my fourth OpenStack summit. Hi, I'm Laurent Lachal from Oven. As you can hear, I'm French. But I've been living in the UK for about 24 years. I'm hardly French any longer, I'm afraid. And within Oven, I'm part of the software team. And I cover cloud. I've been covering cloud for about four years. And OpenStack, obviously, is part of it. Bonjour. That's my French. I'm Sean Michael Kerner. I'm a senior editor at Quinn Street Enterprise. I write for multiple publications there. Shortlist is eWeek, Datamation, Linux Planet, Linux Today, Enterprise Networking Planet, Server Watch, Enterprise Networking Planet, list goes on and on. But those are the big ones. First time I heard about OpenStack was when Chris Kemp called me four years ago ahead of Oskon and told me he had something to talk about other than my theories on aliens living in Area 51. Thank you for the introductions. So I wanted to start by talking about the recent acquisitions in the market. There's been a ton of acquisition. There's been a lot of funding, which can send a little bit of a mixed signal, right, as to where the market's headed. From your perspective, I'd like to know, what do you think this means for not only the community, but for the vendors themselves? And more importantly, their customers. And Sean, I thought I'd kick it off with you. Sure. There's a lot of money. But the money is not necessarily directly attributable to the opportunity. I think the money is partially due to the macroeconomic trends. The fact that money is cheap, right? There's no returns in the general public market. So venture cap is big. So there's a lot of money there. So that's one. In 2008, I think there was just as much opportunity, but no investment. So that's just a function. In terms of risk to potential users, I think the risks are huge. Because when companies get acquired, things tend to go bad. When companies run out of money, they tend to go bad. But that's the opportunity with OpenStack, when it's really pure play open source as opposed to open core. The theory with open source, of course, is there's no vendor lock-in. Reality is, of course, that plenty open source solutions are not. They're actually open core, which means that even if you stick with one and you get locked out, because your vendor goes out of business, it gets acquired and gets shut down, you still may have a risk. So that's my short answer. Well, it's the reflection of the maturation of the OpenStack market, this consolidation. It's as much about market opportunity for incumbents as it is about gaining the right skills, because there's a lot of interest in OpenStack, but not that many people really understanding inner working of the technology. So that's what people acquire, what they acquire a new company, not just the technology. In terms of impact on users, I would say at this moment in time of the market, it's not that much, because first there's not that many enterprise users. And also what they do at the moment is a lot of proof of concept rather than production. So for production is a bit more of a challenge. At the moment, a lot of people talk about portability between these distributions, but it's not really real. But yeah, I think at the moment for productions, that may be a little bit of an issue, but for the market at large less so. And I'll try to, I agree with what the two gentlemen said. Obviously, the M&A and the investment clearly shows a validation for OpenStack, but it also could cause some troubles down the line. So instead of having all these small companies being a counterbalance to the big vendors out there, when the likes of MetaCloud and Inovance and Cloud Scaling go away, and there are some big vendors that kind of control the agenda now, so it'll be interesting to see how it evolves and how the roadmap continues and how the different projects continue with some of the larger companies kind of dominating now. So it's a good thing in general that there's M&A, because people are interested, but it could be a double-edged sword at some point. So I come at this from a slightly different perspective, simply because the more M&A there is, the more funding there is, the more I have to write about. So it's great, it's great for us, right? We don't worry so much about the users who, we see a lot with startups. When startups get acquired, they get shut down. It's bad for the user, obviously. But with OpenStack being open source, I wouldn't worry about quite as much. From that, it's just, as a writer, we look at it very differently, so. But are there worries? Yes, but I think the money, as you said, the money's there right now. It's easy to get funding. It's easy to get funding in Silicon Valley, especially right now. So the hype is there, the hype around OpenStack is there, containers, all that. It's just working really well if you want money right now for one of these products, it's there. Yeah, I think everything is set. I agree with the opinions here. And yeah, it's definitely, it's a threat if the smaller companies are getting acquired after a time or very early. For the users, definitely. But because when you're running a service and we had much of them also in Germany, who knows? And all of a sudden it says, yeah, okay, now our service is shut down in several months. And yeah, get out of it, please. We are giving you the opportunity to get your data, but then you are on your own. And this happened with more the non-OpenSource-based companies, but of course when it's about OpenStack or general OpenSource technologies, it's much easier for the user at the end. Thank you. I'm gonna take a little bit of a pivot. Laurent, you've mentioned that there's a lack of an enterprise adoption with OpenStack. I was curious, especially as an analyst speaking to customers or potential vendors that are interested in using OpenStack or leveraging it, where do you see the market now? Who is using OpenStack? And do you think the enterprise will eventually adopt? Absolutely. I mean, the fact that the enterprise adoption is not that large at the moment, it's absolutely fine. It's not a problem, okay? It's not a criticism. There's two ways of adoption. The first wave are people who are large and have the skill and capabilities to deal with the complexity of upstream code. So you have, well, public cloud providers, internet and DreamHost. You have large scientific bodies, and CERN is a good example of that. You have the IT industry in general. I mean, the largest deal in OpenStack so far has been the deal that Ericsson did with Mirantis, 30 million earlier on this year. And this market is still growing, so it's a good opportunity. And then there is what I would call the second wave or that adoption, which is enterprise. And the second wave of adoption is driven by three things. First is driven by the efforts of the foundation to make OpenStack more usable, more interoperable, more consistent across the various projects. It's driven by the efforts of large enterprise-centric vendors, the so-called incumbents. And again, we're talking about M&A as if it was a bad thing, but I'm sorry. Red Hat and VMware and Oracle and there's the incumbent. Yes, some of their OpenStack moves, some maybe debatable in terms of the detail, but they have the mean to make OpenStack a reality in the enterprise world. And then also the efforts of public cloud vendors using OpenStack. And from that point of view, actually, at the moment, enterprise adoption is on the public cloud side. I mean, Rackspace has plenty of small enterprises using OpenStack. They don't particularly care or know that they are using OpenStack, but that's what they are doing. Do the rest of you agree or does someone have a different perspective they'd like to share? I'll just take the counterpoint just because I would argue that the reason why there may be slow adoption is because not everybody needs a cloud. It's the simple truth. Reality is plenty of enterprises are doing just fine with basic virtualization. What's the difference between virtualization in the cloud, metered and service-based consumption? So for certain small, medium-sized enterprises, do they actually need service and metering? No. On the open-source side, they can run over to Red Hat Enterprise Virtualization or Zen Server or Zen or some other combination. But to actually talk about enterprise adoption of OpenStack, I think the larger question is enterprise adoption of private cloud. That's the larger question, whether enterprises on the whole actually need private cloud or not. Some do, some don't, and that's where I see the adoption happening. But the question needs to be asked in some enterprises whether they actually need it. The only other thing I'd add in is that over time, I think the line between server virtualization and private cloud will blur entirely. Most vendors that pitch me actually don't know the difference today anyway, but there is a fundamental shift and a difference there. So at 451, we have a service called the Infopro, and we talk to enterprises all the time, and we asked them specifically about where they're putting workloads now and in the future. And two years from now, half the workloads in enterprise are still in a non-cloud environment. Public cloud is a very small piece. The growing area is private cloud, and that's where something like OpenStack fits. While it was originally conceived to be an alternative to AWS, it's increasingly seen as an alternative to proprietary platforms used for private cloud. But the adoption rate for enterprises is small in comparison of the overall market, but clearly growing. But the opportunity is in private cloud. So I agree with Joan on that. Maybe just from the startup perspective, the side of the business I see most of the time is I'm seeing more and more companies, small startups who don't want to lock themselves into AWS and are looking for alternatives, and maybe even thinking about running their own servers again. It seems like a switch from the last couple of years. I think there's an opportunity there as well outside of the enterprise for something like OpenStack. Absolutely, and the question is if enterprises should really care about then OpenStack, or if they are just thinking about, okay, I want to have a private cloud, or I want to have a managed private cloud, and they actually don't care about what's underneath. So if it's an OpenStack, if it's an VM, where it's a for Microsoft, as long as the use cases are being able to be developed on this. So of course OpenStack is a very nice approach because it's open, but when you're talking to some enterprise CIOs, they say, okay, yeah, we know OpenStack, we like OpenStack, but in the end, we don't care so much on the infrastructure level or on the management level. So it's really about which kind of deployment model you want first and then it's into the side, okay, if the infrastructure layer we are talking about today is really very important. And I want to say that OpenStack has no meaning, of course, but you also have to think about in this circle. Wonderful, thank you. So next question's a little bit different, but we often hear OpenStack referred to as the Linux of cloud, and I was curious if you think this is a helpful analog, or if it's fundamentally flawed, Al, why don't we start with you? Okay, so they say go out on a limb because that's where the fruit is. So a lot of people conveniently like to make the parallel with Linux and OpenStack, but I actually wonder if we're still in the Unix stage, and as more and more companies make more money and are looking at capitalism, we potentially see a vendor drive off and create the Linux of OpenStack, and maybe we're still in the Unix stage. Our market monitor service, we track the size of the market, and it's a little under a billion dollars in 2014, and 1.7 in 2016. The biggest growth is in that distro area. So it'll be interesting to see as the time ticks away if some of these larger providers don't like what the rest of the group has to say and kind of says, you know what, hey, we're gonna go it alone, and perhaps that becomes the Linux. So not everybody's saying that, but it's definitely a possibility. If anybody wants to totally disagree with that. Yeah, I think it's 100% true that OpenStack is the Linux of the cloud, and you can really see that also the foundation or the OpenStack community has learned from how Linux has developed over years. So in the beginning, also Linux was a very small project. Not a lot of people cared about. Everything was made out of the trunk or you have to deal on your own. I can remember when my first Linux distribution was from Susie, because I'm from Germany, and it was really hard for me because I also have to deal with the kernel and everything, and the graphic card doesn't work and everything like this. And I think that it's good to see that OpenStack is really professional. So you can also say OpenStack is a professional Linux of the cloud because say the foundation, the vendors who are involved in this community really know how to deal with it and to make it professional and to... And also, yeah, the marketing around helps to have a bigger adoption in a very short timeframe. When you're thinking about that OpenStack is only four years now, I think in the past we would not have a Linux summit in this size now here. I'll just give you both sides. I think Linux is exactly the same as OpenStack and then OpenStack is completely different than Linux and I'll explain. You cannot run OpenStack anywhere else except for Linux, except for Solaris and that's just a small kernel difference, but the neutral chain is exactly the same and then there's detrace and maybe you can get detrace on Linux. You cannot run OpenStack on Windows and there's a very good reason for it because OpenStack is Linux. The core drivers that enable Nova Compute, Swift, Neutron are all Linux kernel space drivers. We're not even talking user space. Then there's user space. So the two are the same. Why they're completely different is Linux was born from a scratch from Mr. Torvalds. It was sitting in Helsinki. OpenStack was born by vendors and then gets pushed down. It was two vendors, I'll call NASA vendor and Rackspace, it starts with the vendors and then it gets pushed down. So the adoption curve is completely different and it's much faster because of the vendors pushing down. So it's a little bit different. Though now there are some grassroots effort, you don't see any community OpenStack distribution today that isn't backed by a vendor. In Linux, it's exactly the other way around. There's vastly more community distributions and then there are the enterprises on top. So that's how they're different and the same. Renee? Nope, they're all? Yep. I just want to add something and this is also Chris's opinion is that we see that OpenStack is actually the starting a new wave of open source technologies in the enterprise. So Linux really make it into the enterprise and now we have OpenStack with coming but also things like Dover for example, which actually makes the operating system like Windows. Also Microsoft is now developing an API so that you can use Dover on the Windows infrastructure. So now we see a new wave coming into the enterprises into the data centers like Measles, like Dover and OpenStack is just the beginning. Dover needs C groups and namespaces both are which are Linux native, which is why Microsoft couldn't just say magically we support Dover because it's not native. So they have to build those hooks in Windows server and then hopefully theoretically they'll get it but it's also Linux first. Yeah, I agree that yeah, it depends on Linux. I agree that it's a unix from a distribution point of view. For me, it's not that helpful on one hand and it's helpful on the other to compare the two. It's not that helpful because Linux is an operating system is a well understood type of software. They are clear standards and OpenStack is an infrastructure service platform. The concept of a service platform is not that clear. There's no standard. So the two are very, very different in terms of nature. It's helpful because it helps people understand OpenStack because there's a lot of people out there who don't make the difference between the upstream code, community code, as well as commercial distribution code. And so comparing that with Linux and the Linux upstream code, the Fedora, open source communities and projects and the Red Hat Linux project helps people understand. And in terms of yes, in terms of maturity, I agree with you that the OpenStack is dramatically much more successful than Linux was at the beginning. And that's a reflection of the interest not just in open source, but in infrastructure as a service and the two concerns of vendors, how to compete with Amazon Web Services and how to compete with VMware. So I agree. For me, it's a good way of describing the project. I think to an audience that's interested but not specialized. It's just a good way of getting a conversation started about it and to explain it just on those terms that people are familiar with because everybody knows how Linux works and Linux Foundation, at least most people do. So it just helps to describe it that way, I think. And then we can always discuss the nuances of it. Great, thank you. So the next question I have is, you mentioned where we are now is obviously compared to where we were with Linux so many years ago. We've moved more quickly. There's more adoption, but I'm just curious. It has only been four years. Do you think it's lived up to the hype? Where do we need to go from here? What are you hearing that makes you kind of laugh and promises you're hearing or supposedly things that are happening? And what do you think is more exciting than you expected especially given those of you who've been writing and involved in this since the beginning? So maybe Sean or Al, you wanna start with that one and then I'd love to get the newer perspective from those of you who have just begun. Yeah, early on there was no hype. Well, all the hype was driven just because of NASA's involvement but it took actually a year, two years till it actually built the momentum. The early stage where Rackspace and NASA got on stage at Oskar and Portland was just though this is something interesting but there wasn't that early indication that it was gonna grow into something so massive so quickly. I think that personally, that took me a little bit by surprise. First time I met Jonathan Price was about a year after the initial startup and that's when I knew there was something more. His early target, which was one of the earliest stories I wrote back in 2010, was to be sort of like Apache, do you think Apache HDTPD? At one point before InginX, that was like 90% of the web just because everybody could use it, it's there. That's what I think the promise is on the private cloud side is that it's just something that everybody uses until something else comes along. Yes, I was surprised by the momentum that quickly built and good job that, or Rackspace, I think Rackspace did a good job and I think that the foundation that took over from Rackspace is also doing a rather good job at building that momentum. The problem I think in terms of OpenStack is I still have a lot of people who simply can't understand what the hell is OpenStack because OpenStack is such various things, so many things. First, it's a marvelously successful brand and I think that's the core success of OpenStack is as a brand. It's got a huge mind share, not yet the market share, but from a mind share point of view, it's Kudos, you're a marketing person, I'm sure you're jealous of that appeal that the brand has. It's also a project, it's a community. As a project, it's a variety of components. Each of them has its own story to tell and a complex story at that. It's an evolving terminology as well. It started with core component, then you have now core components, related components, incubated components, and then components waiting in the wing in Stackforge. And then the notion of core has now completely changed to moving from the core technology components to what it means to be a core implementation. So the notion of core has moved from upstream code to downstream distribution. And that's just the beginning. I could go on and on on what the hell is OpenStack and all the various facets. And so no wonder that people who just want to familiarize themselves with OpenStack have problem because of all this. And I forgot to mention, you have the technology components, the storage, the compute. You also have the support components, the documentation, the release, the management, which are not, which don't get enough light and I think as important for the wellbeing and long-term momentum and success of OpenStack as the technology components. So yeah, it's very complex. And I think that's the last thing I said. I think the vendors are not doing a good enough job to explain to the market what it is, what OpenStack is. They are too busy papering over the crack of OpenStack in order to sell whatever they have to sell. You know, I want you to ask that question, but you bring up marketing and the hype surrounding it. And do you think it's because certain vendors, I'm not gonna say myself included it, no, but we're kind of so in it and we're drinking the Kool-Aid and it's been four years. Do you think that's the reason that we've skipped past the education phase in a way? Because we feel like, well, who doesn't know it? Does that make sense? And so maybe we need to kind of take a step back as a community and I'm just curious from your perspective on, Al, if you wanna start and then, you know. It's clearly moved fast in four years, just the growth of the community and the amount of investment, but it's still as many people would agree, it's difficult to consume. And I kind of go back to like the, my oatmeal analogy in the old days of oatmeal, your mom had that cardboard canister and she took some out and put it in the pot and if you didn't have enough water, it was too lumpy and it was a pain in the butt to clean the pot. And what OpenStack needs to get to is instant oatmeal where you rip off the top, everything's in there, all the ingredients are in there, you just add a little bit of water and you're good to go. And it needs to be the variety pack of instant oatmeal. So if you want the big data version, you know, you rip open the packing and you make that if you need a private cloud for web hosting, like it needs to be that simple to consume in order for enterprise to have wider adoption. That's a tough one, right? One moment, do you have a question you wanna ask or it really needs to be something? I just sat through a theater presentation, actually, from your company where they were basically presenting instant oatmeal. The question I walked away with is, okay, but what do I get for $4,000 a year per server with instant oatmeal? Somebody should be able to explain that. I'm not from a vendor, I'm from the analyst community and I agree that the vendors need to explain exactly. If you walked away from a vendor panel and don't know that answer, perhaps they didn't do a very good job explaining it. But, you know, there's different ways that people are consuming open stack. There's the enterprises who think they have the resources and they'll go download the code for free and fight through it and turn something up and that's kind of the old school oatmeal way. There is the distros and the product providers which is the fastest growing business model who will sell you a product and you will still go and put that on your own hardware and turn it up yourself and that's kind of what the, you know, red hats and the suces, the canonicals, marantis and pistons and those will do for you and you'll pay a new annuity to do that. And then there's ones who will, you know, just offer you a service, you know, whether it's MetaCloud or Rackspace or in internet and the likes that do blue box that do private and public cloud. So it really depends on how you wanna consume it and how much you wanna take on yourself versus outsourcing. Okay, well, just one thing in the interest of time and I'd like to leave room for questions and I'd be happy to answer your question as it relates to piston after this panel but I love the gentleman on the end to also have an opportunity to respond to that question if interested, so. Just wanted to pick up the education part that you mentioned earlier on. I do think that has, that was skipped because people have been so familiar with it in the community but people outside of the community are just now learning about the project really in the last year or two, just the last year really for many, many people. And it's so complex and then for me trying to learn about it, I met with a lot of the executive team at Oskon this year for the first time and it's really where I started picking up things but then you start reading about it and there's code names and there's real names and it's just, it's a really complex project. It's not easy to sum up and you can call it the Linux of cloud or something like that but there's a definite need for more education I think. Yeah, I totally agree, education is still important and well it depends on the provider or the vendor if you need to educate an open stack. For example, if you're a piston cloud, if you are a blue box or a rec space also, the question is if you really need to educate the people because it's open stack, it's running. So you have a data center as a service or an open stack as a service but the people only need to educate on how to use the APIs and not what this whole stack really is. This is a question you have to answer as a provider. Of course when you are a distribution provider like a Myrentus or a Reddit or a Suzy, it makes sense to have a story around because the people are really using your distribution and they should know what they have on their infrastructure. And yeah, I think these two questions are very important for the vendors. And regarding the hype, I think the hype is over. The people actually know what open stack is so the important people you want to address and yeah, they see it like a real option to infrastructure solutions like from Microsoft or from VMware and you can really say that open stack, we always hear that open source is eating the software world. Well, I think it's open stack is eating the software license-based world. So companies like Microsoft or VMware really will struggle in the future and we are talking to so many enterprises who say okay, we are now on a VMware infrastructure and we don't can handle these subscription change anymore because it's getting expenses over the years and we really need to find an option to, for example, change to a KVM or a Xenhypervisor, yeah. Sean, do you have a perspective? I know being on the ends of the panel, I can. And then I know, Laurent, you do too. Just on the question of what you get from the vendors. In open source, and I learned this very early on the first time I met Bob Young in Toronto who was the founder of Red Hat years ago. And he was just starting to sell the stuff and I was using it for free. I said, well, why would I pay you? And he told me, well, you're not paying for the bits. So when you consume any open source software, you are not paying for the software. If any vendor tells you that you are, they don't understand open source and just walk away. Or they're open core and they're not really open source. What you're paying for is sometimes the glue. For example, if you had open stack and you wanted to download from trunk and build your own, which I tried to do once, I don't recommend it, it's a path to insanity and my sanity is questionable to begin with. So there's that, but you're paying for glue maybe and service and support. But in open source, the bits are open and free. Yes, agree. So you get integration, you get add-ons, you get support, you get implementation services, you get a lot of stuff. There's a lot of stuff that you can wrap around open stack. One thing I wanted to add in terms of the nature of open stack is that it's still evolving. I mean, at the moment, what I see in the market is a continuum between companies that define open stack as an API layer on top of their own technology. That's the VMware and Oracle view of the world. That's one side. On the other side, you have people form open stack is both the API and the implementation of the services underneath. And it's the continuum. There's no right or wrong, although there's a lot of debates in the right or wrong of that particular continuum, but there's no right or wrong. It's just about what is relevant to a particular market. The API approach is relevant to a particular market, to more legacy, to more traditional companies. The implementation is more, it's more than to a more startup, open source-minded type of audience. At the moment, this continuum is being defined as the community talks about it and as the Linux, sorry, the Open Stack Foundation works around this notion of deathcore because it's unclear at the moment whether this deathcore project will stop the API, whether and to what extent, it will limit the API approach versus the implementation approach. And that's the moment, that's the real key question and it's an evolving question. There's no answer to that question and it's okay that there's no answer. It simply means that it's an open source project, it's built on consensus and that consensus will take time to build. Great, well, we're almost out of time so Renee, I know you had one thought and then I'd like to open it up to the audience in the last couple of minutes we have to see if you have any questions. Yeah, thank you, just a short addition. The lack of, yeah, we are talking about education, I guess it's mostly not the requirements of the, it's not the vendors or the providers should take care about, it's more the ecosystem around. So when you're for example taking a look on the OpenStack marketplace, we don't see a lot of system integrators or consultancies around and maybe it's a part of the vendors and of the providers to educate first the ecosystem, the partner network and that the partner network is running out, spreading out and tells a story out there because in the end, these are the people, the consultancies who are creating these massive infrastructure based on OpenStack or in the enterprise infrastructure on-prem or in the data centers. Great, do we have any questions from the audience? I know we have a minute or two left. Do you wanna step up to the microphone? Thanks, very interesting discussion on the educational element of this in the comparison to Linux. I think one of the problems that the community has is that because of the rapid evolution of OpenStack, the business value messaging hasn't really caught up because you can talk about Linux and you can use that analogy, but I would argue that many line of executives in businesses wouldn't get what the value would be to them as a line of business person. So I think we all are kind of jointly responsible for doing, I think a poor job of messaging the business value of OpenStack and I just was curious to see the comments on that. Yeah, I can totally agree. Our research study on the OpenStack, okay it was only in the German market said that we are asked for the pro and cons of OpenStack and one of the cons was that there are not a lot of reference projects out there to see how you can use OpenStack in your infrastructure for your use cases. So yeah, it's right. So we need more of these stories we hear today at the keynote, how Expedia does it or how CERN does it. It's really important, yeah, definitely. I disagree a little bit. I don't think it's up to OpenStack to educate necessarily though they do. I think the way that OpenStack is adopted is users up. So for example, CERN was not educated by OpenStack first. They had a problem, they look for a solution, they go same with the BMW keynote the other day he had a trouble. His guys were coming, giving him reports, he goes, tries it. The challenge for line of business is simple. I have applications, I need to deploy, I need to do dev test quickly, how do I do that? Figure out the answer, if OpenStack is the correct answer that's the end of the conversation. As technologists, trust me, I love to talk about the technology, it's not the point. It's an enabler for business to the agile deploy applications quickly at the lowest possible cost and the highest possible degree of efficiency. The number one reason we hear from enterprises as to why they're even considering it is cost. So they're looking to either take workloads that are getting more expensive in AWS and doing it in a private cloud environment or with another service provider or they have a large embedded base of a proprietary platform and they're looking to do it without having to pay that annuity license cost but what the service providers and the distribution providers need to do is show that cost-benefit analysis have a calculator that shows you if you're spending this much in AWS we can save you X by moving it here or by switching off of your proprietary platform to this distro instead you can save Y. I think they'd need to articulate that a little better than they are. Just quickly I know there's another session coming up and I'm sure we have a million questions so if you gentlemen wouldn't mind maybe we can go outside for a few minutes after and address some of the questions that maybe didn't get answered. I won't have respect for the next session which is starting pretty soon. So if for those of you I apologize that we didn't have enough time but it means lively conversation so thank you for attending and thank you all for participating and again if you have additional questions that weren't addressed please feel free to meet them outside and speak with them.