 All right, when we get started, first of all, thank you everyone for coming to this panel's discussion, discussing scaling out OpenStack Clouds and Enterprise. We've got a great group of panelists who will be talking about their experience and what they're seeing today in the industry. So some, by way of introduction, my name is Kenneth Hoy. I'm a technology evangelist at Rackspace and I'll be modeling this panel. If, when we start by just kind of going down the line and quickly give you a name and talk about what you're doing or what you were doing around OpenStack. Hi, my name is Caroline McCrory. I am now Senior Director of Research at GigaOM. I was the Head of Product at Piston Cloud, which was an enterprise OpenStack software company. Yes, hello, I'm Jan Holzer from Red Hat and I do a lot of different things in Red Hat, usually anything around emerging technologies, evangelizing technology, but over the past two years, really focusing on OpenStack and helping customers to deploy OpenStack and also supporting what we call an early adopter program around OpenStack, not just with customers, but also partners. Great, I'm Jesse Proudman. Up until last week, I was the founder and CEO of BlueBox. I hired a new CEO, announced Monday this week from Tier 3. Now I'm founder and CTO of BlueBox. We do on-demand hosted private cloud. Business has been around for 11 years. Sorry, I don't have a microphone to share one with you. So my name is Boris Ransky. I am Co-Founder and Chief Marketing Officer at Morantis. So based on my title, I think it's easy to figure out that I run marketing. Also, our training team reports into me and I'm also responsible for kind of personally creating some of the strategic partner relationships between Morantis and OpenStack ecosystem partners. For those of you that don't know Morantis, we are the pure play OpenStack guys who have three lines of business, subscription business, for Morantis OpenStack distribution that is similar to many other distribution options available out there, training and services. So that's a summary. Thank you, Boris. My name is Manju Ramanathpura. I work for Hitachi Data Systems. I'm a CTO with Hitachi. My primary role in the past year and a half has been to really define overall strategy for Hitachi around OpenStack. Hitachi itself plays very strong in the storage business as well as the converged infrastructure business. When we look at OpenStack, essentially, we are looking at really building a cloud portfolio built on OpenStack and that's the group that I'm working with to build the strategy. Okay, thank you all. So just in terms of the format, I've got a few questions. I'm going to throw out to the panel. They can all, everyone will choose to answer those questions or you can choose to only answer some of them, but we'd like to have some back and forth obviously between all of you. And you notice that we have a mic in the middle. So hopefully sometime during this session, we're going to take some audience questions. If you do have those questions, my suggestion is you kind of start lining up in the middle so that we know that you have questions to ask and that way everyone can hear your questions and it'll get captured for the YouTube video. So the first thing I want to ask though is there's been a lot of conversations in the past year about is OpenStack actually ready for the enterprise? And in fact, if enterprises are even interested or whether they think it's ready for the enterprise. So kind of a very general question based on what you're seeing the market to feel today. In fact, what is the state of enterprise adoption for OpenStack? Anyone can take this question. I can start I guess before I have to give away my microphone so I'll take the opportunity to use the microphone. So the first part of the question about whether or not OpenStack is ready for the enterprise and I don't think that there is a kind of, yes or no type of answer. I don't think that anything is ever completely ready for anything and I confirm that indeed it is not an absolutely perfect product and there are challenges with OpenStack as are challenges with any software product but as far as the adoption is concerned I think that this year it is clearly starting to take off and I'm kind of glad to see that about two years ago I've kind of approximated the path for adoption of OpenStack by various segments and the first kind of line of adoption was really primarily around the infrastructure vendors just kind of writing drivers for OpenStack or playing and tinkering with it so it was just complete vendor fest and then the next path was really the very hardcore tech savvy guys primarily in SAS Web 2.0 segment that started adopting it and two years ago I said that 2014 I think is going to be the year when enterprises are gonna start really kind of taking OpenStack seriously and based on what we're seeing from our customer base as well as based on what we're seeing from the vendor announcements, the traditional vendor enterprise vendor announcements is absolutely the case that enterprises are adopting OpenStack today so I can say that we have customers, enterprise customers that are adopting it but I think that another interesting proof point is the fact that just recently all the three large enterprise vendors, IBM, Cisco, and HP made these announcements that were committing a billion dollars to OpenStack and the reason why they're doing it is because they're seeing their enterprise customers ask them to have some sort of OpenStack strategy and I think just to be virtue of them announcing it is a proof point of enterprises kind of tracking with OpenStack. Yes sir, I would agree, I would have said yeah, first part is lower case yes and second part is upper case yes. So clearly OpenStack has a huge brand, has a huge awareness in everybody's customer base, every customer wants to talk about OpenStack. If you don't talk about OpenStack today they just go to another vendor and talk to them about OpenStack but I think from implementation perspective I would probably agree with everybody on the panel. I always say it's still a box of razor blades today for many people. It's not really a simple to install like say a VMware perception, just come install, run it. It requires a lot of hand holding and we run still through a lot of education as well. You shouldn't just take your enterprise workload, rehost it and run OpenStack but actually do something meaningful with it. So I think right now we see a lot of customers who have a lot of interest but the usual suspects say on the FSI sector, internet companies, as you said, Web 2.0 who actually do meaningful workloads and run meaningful installations because in all we've seen from the user survey there are not yet a huge amount of very large scale OpenStack installations. There are a lot of medium sized installations and the large scale ones, they're often then trapped in prior releases because there have been a lot of challenges in upgrading, stability, all of those things but I think we get there over time absolutely and we've made huge strides in Havana, the Ice House and to know around upgrades, around stability, I think networking is still a fairly complex topic and I think it also exposes an issue around skilling. I mean, a lot of customers were used to, I just have regulars as admins and now suddenly I really need like a network architect, a storage architect, I really need to even consider application architecture for those deployments. I think that from my perspective I'm finding that there's a terminology issue. Since a lot of people are using VMware, they have in their minds, HA is the way VMware does HA or FT is the way VMware does FT. So a lot of the customers that I've come across even in the research space now, they're all wondering, I want to be able to deploy OpenStack, like you said, Jan, very easily, like the way if I could just buy vSphere and just put it in and I believe that OpenStack is now maturing to be able to get to that perspective. There's more people that are contributing the code, there's more people who understand enterprise operations of IT and it's bringing that type of DNA into the OpenStack ecosystem and putting in that code and having people understand how to actually do those sort of deployments. I'm a big fan of this term service catalog. I talk about it a lot. I think OpenStack has a ways to go from a service catalog development perspective before but really see widespread adoption. So I use this analogy of a tree. You've got a strong trunk, core services, compute, networking, storage. For the most part, we've got the kinks worked out there with the exception of networking is everybody's joking about. Then off the branches, you've got additional services, database, caching, queuing, load balancing, these types of technologies and to me the OpenStack technology has not evolved the additional services really necessary to architect modern web applications on top of the platform effectively today. So much of the adoption that we see are implementations where it's DevTest, I guess I'm getting your workloads now, but they're DevTest workloads or there are other components that aren't relying on a full suite of services that you may get out of a true public cloud. So I think, yes, there's adoption, the service catalog needs continued development before we'll be able to see that rearchitecture of applications truly occur. So, I'm sorry. Yeah, I was just going to say, I think pretty much in the enterprise, one other thing that I'm seeing is lot of the enterprises are looking for a dual vendor strategy and OpenStack sort of initially became as the second alternate option, but as Boris also mentioned, you're seeing lot of these large vendors doubling down their investment with OpenStack and the reason is the customers are driving that and from a customer side, what's really happening is OpenStack is actually becoming the primary vendor and then there is a legacy applications, there is a legacy investment already done with whatever the alternate solutions are, they're not ripping that apart, but the new age applications, if you talk about Hadoop type of workload, those workloads, they're not even thinking twice, will I run it on anything other than OpenStack? So that's the trend that's happening in the enterprise, right? And I think as a community, I would also think we all have a lot of responsibility on our shoulder to help spread this message as well. It's been great when we look at all these keynote speeches, we had Wells Fargo coming in, we had Ericsson talking about it, AT&T talking about it, we're having really, really large customers who are deploying OpenStack in a production environment coming and talking. We need to build on that and actually helps do a service to our community, OpenStack community to make our customers, other customers also feel comfortable and make analysts feel comfortable to make recommendations to their customers, yes, OpenStack is ready. Don't worry, it's not yet ready, wait for another five years or something like that. So several of you have mentioned VMware in the context of your answer. So if you grant that probably 100% enterprises today are using VMware in some fashion, I guess the question I have is, why are enterprises even looking at OpenStack? For example, I've talked to a couple of enterprise customers who've said, I have a private cloud, I've had it for four years now, see, it's running right there on a V block or on a flex spot. So how would you answer that? Why do you see enterprises looking at OpenStack today? And then kind of a later question is, what are the use cases for running something on OpenStack versus, say, their VMware or their Hyper-V environment? I'll just take it and then pass it back, sorry. From my side, there are probably three major reasons why enterprises are looking for OpenStack. One is flexibility, having the ability to hack the code, make the sort of unique data center differentiation for their data center. These are the large vendors, right? They want to have some differentiation. They don't want to wait for any proprietary vendor to make the changes and take a year or two years to deploy it in their data center. That flexibility, they absolutely love it and that's the benefit of OpenSource. And the second thing is, there's also the advantage of really building on top of the community as opposed to relying on a one vendor solution, right? More and more customers want to actually have that no lock-in to one specific vendor as the cloud really is evolving. There are so many things that are converging together in the software world and as the convergence happens, there is more and more fear of, oh, what if I get locked into one or two vendors? And then really the third one I think is the cost as well. I don't think we should underestimate the benefit of very low software license fee or a service fee attached with OpenStack compared to alternate solutions. So I think it's entirely about that service catalog. I'm gonna bring it back. So Jonathan Murray from Warner Music has a really great presentation about the Composable Enterprise. Let's think about where are application developers taking enterprise applications in the next 15 years? And it's not about traditional application development that sits on individual servers. It's about interacting with hundreds or thousands of different API endpoints and being able to consume self-service services from a cloud. VMware is not a cloud. Everybody has private virtualization is what they have. And when people think cloud, what they want is the flexibility of something like an Amazon web services. So how can I get that catalog available to me either in my data center and a private capacity? That's where I think people choose something like OpenStack as a technology platform. So I should probably not talk about OpenStack and VMware, but I'll take a stab at it anyway. So I think that both of you guys have touched on a very important theme. And that theme is the aspect of flexibility of the components that plug into OpenStack. If you look at the survey that the foundation has put out to all of the OpenStack users, the flexibility of the underlying components is inside the top five business drivers for OpenStack adoption today. And another interesting point is that the other four are really kind of generic kind of cloud drivers, like increased agility, better engineering velocity, et cetera, so the only kind of unique business driver to OpenStack is that flexibility of underlying components. Now, on to VMware, I think that VMware now has a very correct understanding of what OpenStack is and has a very good positioning of how they play with OpenStack. And what VMware is saying, and I completely agree with it, is that basically OpenStack is this glue that helps you glue together fairly heterogeneous application infrastructure. And VMware is just one of the innovators specifically around virtualization, software defined storage, et cetera. They're kind of the vendor that provides lower level components primarily for the IT world as we're transitioning from kind of the single known mentality to the completely distributed mentality. And I think this is an extremely important concept because if you track back a little bit to history and see how kind of Linux and Microsoft were born, both of them were kind of a child of effectively hardware, software desegregation that has happened in the computing industry. In the early 80s, to seed the industry, the big guys were shipping mainframes as a kind of single integrated solution, hardware, software on top, and you need to do that to pierce the market. But then as the market grew, there's been a lot of innovation, particularly in the hardware area after x86 was born, and no single vendor was able to actually keep up with that innovation. So ultimately, there's been an alternative. Alternative became Microsoft and then eventually Linux. And the reason why everybody wanted it is because these were the fabrics that could pull together all of the innovations that were happening in an increasingly fragmented kind of x86 ecosystem. And what we're seeing today is we're moving towards the kind of completely distributed paradigm. There's yet another desegregation that is happening, but now this desegregation is happening not between just hardware and software, but desegregation in the layer of vendors providing fabrics for the distributed world. And OpenStack's emerging is this kind of glue that puts together all of the innovations in the networking and software and virtualization and exposes them to everybody through a single open fabric. And that's really what I believe is kind of the key value behind OpenStack. Yeah, so I actually share the same perspectives as Manju. Most of the large enterprises that I've come across have actually been interested in wanting to invest in OpenStack because it's a lower cost point. And they're looking at either building completely different types of workloads to Jesse's point, to put onto OpenStack because of the flexibility that the OpenStack framework brings, but also for them to actually think about how they've put workloads onto VMware and which of those workloads may not be suitable to stay on the VMware ecosystem. Not that they're gonna entirely dump VMware and put everything on OpenStack. They're looking at what's a good mix for the strategies that they have. Yeah, so I think that's something we see a lot is clearly the perception is I get these huge savings, I get the low cost alternative. And that's not necessarily true today, where you can't just simply rip out VMware, plug in OpenStack and would do OpenStack at this service as well. Sure, you can plug in VMware underneath OpenStack, we can argue all day whether that's useful or not. But I think I agree with Jesse and everybody else around it enables customers to take advantage of much more rapid innovation around services, also come to a much more common set of infrastructure APIs endpoints rather than just having a single vendor dictating, for example, the APIs, how they can develop their applications, how they run applications and ultimately get to an easy style infrastructure internally, externally, but really take advantage of a much, much broader ecosystem. But then again, it brings a lot of challenges to all of us all around certification, support for all of this complexity, but ultimately it gives customers a much, much better choice. And all the longer term, obviously a lot of interesting cost opportunities. So who is bringing in OpenStack into an enterprise today? Is it the executives, is it lines of business, is it developers, infrastructure guys? Who are the ones in enterprise who are looking at OpenStack and bringing it in? So I think from my side, it actually can be almost all over the board. So if you talk about who brings OpenStack in as a conversation topic, usually it's anybody on the C level. They've read it, they've seen it everywhere. I need to talk about OpenStack. You literally can kick in doors today about talking about OpenStack. Nobody would open the doors talking about virtualization. Barely cloud probably excites people as a topic, but OpenStack, everybody wants to talk about it. But then there are also the folks who actually want to do something meaningful, real with it. And that's where we probably see a lot of the development teams caring about something like OpenStack where they want to build a true pass, deaf ops type environment. That's where OpenStack often comes in as sort of a real implementation opportunity. And then we see a lot of proof of concepts today around sort of from the CTO side, early IT architectures. For example, again, customers who look at the next generation data center infrastructure and how can I leverage OpenStack in this context and how would it make sense to bring OpenStack into my environment and also attract internal customers to actually use it. Not just run a POC with a couple of nodes and then it's idle, but really draw people onto OpenStack and develop applications and then actually broaden its reach with Indone Company. I want to make just one comment. I'd agree with what you all said. But I also want to emphasize on one of the really cool things about OpenStack, which is the decoupling of various data center components, storage management, server management, network management. I was involved with the previous incarnation of private cloud in a box solution with alternate solutions to open stack. One of the pushback I used to see from the customer at that time was, well, in my organization, I have a storage team, server team, network team. You're trying to collapse that team and trying to shake up the organization structure and there was resistance coming from the people who are actually operating the data center. And I don't see that resistance with OpenStack. I have a feeling part of the credit goes to how the OpenStack itself is architected. You still have the ability to have a data center administrator specifically focusing for network and storage and server. They get to do their specific roles and responsibilities and when we were getting the pushback from the operators, it wasn't about job security, it was really about they know that there is something very true value that they add from each of their silos, right? The network silo, storage silo. And if you try to bring all the pieces together, it shakes up the integrity of their operation. And I'm seeing that that's one of the nice thing about OpenStack and you've got even the operators now really embracing OpenStack in addition to the CIO CEOs. Yeah, Matt, I think we're kind of approaching this question in the wrong way. If organizations are talking about bringing OpenStack in because it's OpenStack, it's an early adopter and they're picking it just because of a name or it's the newest technology. What we see in our conversations is that the C-suites and the boards of directors and enterprises are saying to IT ops, look, you need a cloud strategy. Everybody else is doing it, we need a cloud strategy. And on the flip side, we've got engineering saying, look, we need self-service. And today they're going out and they're getting that through whatever means they can't. So I've been credit card with public clouds or other avenues to be able to move at the velocity they're being asked to move. And so IT operations have been stuck in the middle saying, well, I've got this virtualization platform that lets me dole out VMs and that's great. I've got cost advantages of virtualization. But I have no mechanism to actually deliver self-service today. And to do that with that catalog that the engineering teams are looking for. So to me, OpenStack represents a solution to that problem and it has become the de facto solution to that problem from an open source perspective. And that's how it's coming up into organizations where the pilots are not just in the CTO office or they'll really be considered more than just a test case. If we're gonna talk about actual adoption and not experimentation, that's what we've seen those initiatives starting from. Okay, moving ahead. So when I talk to enterprises, one of the most frequent questions I get asked is, they say, why are there so many vendors in OpenStack? Well, and I drill down to find out what that means. A lot of times what they're saying is, there's a lot, they are confused about who they should go with, who should they work with. You know, for example, Oracle just released OpenStack on Solaris. So now you've got another version of OpenStack and that's causing a lot of confusion in the enterprise. So how would you answer that question? Two versions of Oracle OpenStack, one on Solaris and one on Oracle Linux, independent. So two distributions. So the question I have for you, are there too many vendors in the OpenStack ecosystem today? And how should an enterprise, what should an enterprise think about when they try to decide who they should work with? Yeah, I mean, are there too many car manufacturers? It's really, it's about consumption options. And each vendor is bringing something unique with their distribution or with their model of bringing OpenStack into an enterprise. We've got a very different model than Red Hat or Piston or Mantis. So I think building that ecosystem is really about building choice in how we bring OpenStack into these groups. And from an operations perspective, it's really the commercial support that you're going to get. Not all the vendors are going to give you the same level of commercial support. So it's down to what your business requires and what your needs are. And you might choose one vendor over the other, not necessarily just from a technology perspective, but mainly from the commercial support for what you're trying to deliver. So I think that I agree completely that the fact that there's so many vendors on the market today is confusing. And I think it actually really honestly kind of stifles the adoption of OpenStack. But at the same time, it's a new big kind of emerging market. Everybody wants a piece of it, so everybody's in it. And I think that I mentioned that this year is the year of enterprise adoption. I think that also this year, we're going to start seeing some consolidation on the market. And I think we're starting to see it already. So some of the earlier guys are starting to dissolve and there's going to be some people buying some people and it's going to start happening and it's going to happen probably. A lot of it's going to happen this year. So, and as far as which vendor to work with, work with Maranze, of course. I'm obligated to say that. I think I need to object now. No, I think, well, we all want to further OpenStack, obviously, in the enterprise. But as Caroline said also, it also really depends on the customer to some extent on their knowledge level, their expectation, what they need around support, what they want around support. And then we all experience various levels of readiness for OpenStack deployments. We have a lot of customers who want to have the same enterprise model for OpenStack than they have from their prior engagements with vendors like us, for example. So that's where obviously we have an opportunity for them using distribution model with updates, security patches, certification, ecosystem, et cetera, et cetera. And there's also set of customers who are very, very keen on being at the bleeding edge, very, very early adopters, where I think the consumption model with OpenStack is different. You can't do these massive upgrades because they're very disruptive. That's why everybody's stuck on grizzly who actually does something meaningful because it hurts to move forward. And I think longer term, I personally believe, the consumption model, how you consume OpenStack bits will change in the long term where it might move more to like a CI type model where you just get continuous innovation and pick based on features rather than major releases and have big upgrades. And again, the community made big strides around upgrades, already being able to do rolling upgrades, those things. But it took a long time just to get to a defined core, like with Oslo messaging, for example, defining APIs, interfaces, get all the discipline in which we all have been living through in the Linux days for many years and we live through it again. I would say there are no new problems. They're just people who haven't met. So I think we go through the same cycle again here, but I think we all do see a much different velocity with those advanced customers who really want to move forward much, much more aggressive than I think the classic enterprise customer has. I wanna make two points. One is, we talked about why are our customers looking at OpenStack. And that's one of the primary things we highlighted was flexibility. And that is one of the reason why you have a lot of vendors. Different customers have different needs and that evolution is happening right now. I'd agree with Boris that there will be consolidation in this year, next year, but customers are also not sure what exactly they want and not every vendor is really promising the flexibility that customer wants. I think there is that challenge. If you go with me, then you have to work with this, this and this. And that sort of like still in some ways locks them. There's that in terms of how do you choose your favorite vendor? I think the advice to the customer is to make a list of what are the things that you would like to accomplish. There's different vendors who are gonna meet those requirements. So you do have to make a priority. If the cost is a priority, you'll have to make that as your primary reasoning in terms of how you choose your customer. If it's the flexibility, where do you want the flexibility? Do you want the flexibility at the infrastructure level, software level, orchestration level, or the ability to pick and choose different components? You'll have to make those decisions. Then I would say, instead of, yes, I would like you to pick Hitachi data systems, but I would also say in the interest of the community, you'll have to really think through what are the priorities for you. Okay. I have one more prepared question. So if any folks out there have questions they want from the audience that they want to bring to the panel, I would start walking over to the mic. But the last prepared question I have is this. So a lot of you have talked about there are still challenges today with OpenStack, particularly for getting it adopted in the enterprise. So if I could make each of you the benevolent dictator of OpenStack for six months. Oh, dear. Right? What is one or two things you would want to see happen with an OpenStack that you think would make it easier to make it more adaptable for the enterprise? Ladies first. Great. Give it to the woman for this one. If I could be the benevolent dictator of OpenStack for six months, I would want two things primarily. Much better stability and much better high availability features because for true enterprise adoption and if you think about the production workloads that a lot of them have, it needs to be stable and where the stability comes when you need to actually be able to do an upgrade of your nodes and the rolling upgrades between nodes, the application shouldn't suffer. And so high availability, especially when you're competing with something like VMware has been able to do, that is what a lot of them are looking for. And I've come across many people who are in the OpenStack space that just go, oh, they're doing it their way. We're gonna do it our way. We're gonna rewrite the meaning of HA. And I don't think that's the right way to go about it. So now I get to dictate, I guess, for six months. By the time we are done, it's like four years down the road here. So I think, well, you mentioned a couple of really important topics, which actually go into, I think there needs to be in some cases more prescriptiveness around OpenStack, and the deployment types, especially, again, around HA just for core services and all the other services. Because today, yeah, choice is great, but too much choice might also not lead to a very successful implementation. And we all, at the summit here, design them, we all beat networking up plenty. Everybody knows that's a real weak point today and the real trouble point, especially around scaling, also brings the whole topic in around partners and vendors and lock-in and whatnot. So it would be great to really focus and get networking in shape. I think deployment itself is also really important. Unify deployment, get to a point where you actually can easily scale up your environments. What we also have seen recently is a lot of interest from the HPC community to introduce bare metal into OpenStack, which opens up a whole another can of worms around how you do this. Of course, the answer would be ironic, but there would be a lot of discussions to be had on whether that's the right approach, at least today, how it works. So I think, yeah, deployment is really important. Again, being prescriptive around architecture, at least at a foundation level, how can you then easily just scale up your architecture and out, and interoperability between vendors as well. So making interoperability easier, certification easier, so you can, as was said earlier, you can easily match vendors, for example, mix and match components as you want. Jesse, I'll leave you some stuff. Yeah, so I'm gonna take the tree analogy back. So strun, strun, many branches. Defcore and the work there, I think is really critical to define interoperability in that trunk, and then we need to look at the service catalog and compare that to what you can get from true other cloud services. So load balancing is a perfect example. Any modern web application needs load balancing, an effective load balancing to be put into production, and today OpenStack doesn't have a good story there. It needs to be split out as its own project and considered another leaf or another branch on the tree. So it's really an evaluation of, and I actually disagree with you, I think we should not try to emulate VMware. I think we should build an architecture that solves that self-service issue. The customers that we're interfacing with are looking to add this in extension to what they have today, and really develop new workloads, and oh good, now we're gonna get a debate about it. No, no, no, we're not, I'm gonna say it to you. I'm not saying that we have to emulate VMware, but I'm saying that you can't turn around and say VMware is doing HA wrong when it works. Oh, works great. But do it so that, make it so that it's understandable to the people that are already using it, not emulate. Understood. But yeah, trunk and branches, figure out what are the services that are needed to continue to drive application development. So, very short comment, I think that OpenStack now is no longer a product, it's just a community, like an Apache foundation with a bunch of people developing a bunch of stuff, and I'm also a little bit concerned that the dev core stuff that is driven by the board is a little bit over complicated and might take very long time. It is absolutely the right initiative, but the first thing that I would do if I was a benevolent dictator, I would narrow significantly the scope of what OpenStack is and focus just on that because I'm not exactly agreeing with the leaves and branches, I think there's way too many leaves and branches. So now with Trove, we're going into the database management, we're doing queuing as a service, load balancing as a service, everything as a service, you can sprawl the features indefinitely. And that's the problem of the people weren't sprawling the features and splitting things up to have their own piece, but instead just focused on getting the core important components to work with have a much better product. So next time when somebody comes in and says I want to do a database as a service, or this is a service, I'm sorry, this is not OpenStack, go do it somewhere else. If you want to be part of OpenStack, go fix Neutron. So. So. So. So. So. So. So. So I'll make it quick. I think I'd agree with most of the comments Caroline made and Jesse as well as John and Boris. I think me being, you know, looking from my perspective, most of the customers I interact with are the enterprise customers. Caroline made a comment about stability, adding the enterprise grade functionality. So far we've discussed about what do enterprises want and why are they looking at OpenStack? And the next piece of the puzzle is what's missing in OpenStack that enterprises really want. You know, if I listen into some of the keynote speeches, you know, Glenn from Wells Fargo talked about, you know, compliance being one of the things and you know, reason, you know, that's, they have to have certain type of backup mechanism. They have to have certain types of rules and regulations that they need to follow in terms of where their data gets stored, for instance. You know, that's, that's one, like a narrow set of functionality. And Boris was saying, we need to narrow down from various different branches of focus for OpenStack. I do believe that that's a good strategy and by doing that, we will help strengthening the pieces that are now matured and maturing, right? You know, Nova Cinder, Neutron. I do think that they all are missing some of the enterprise features and functionalities. I would probably double down on that and try to, you know, make them ready for some of the gaps that enterprises really are trying to fill right now. Okay, great, thanks. So, this gentleman's very patient. So you could ask a question quickly. We'll get one or two people. Since I've been so patient, I'll also make a comment. Okay. Do you have any other dictators out there? Just underscore, remember, at the end of the day, it's all about business value, despite all the technical deep dives and technical branches and rabbit holes that we have to go down in the community. Now I'll ask my question. So do you have any war stories of any enterprises that have gone down a path of implementing one distribution of OpenStack and have either stopped due to maturity of a particular release of OpenStack or have found that, hey, we really need to jump to a different distribution because of things that are emphasized in that particular vendor's distribution? I have an example. I'm not gonna name the vendors, obviously. I do have an example. It goes back to the flexibility, not having the flexibility to pick and choose the type of hypervisor that the specific customer wanted was a reason to move out of one distribution to another distribution. I'm sorry, we have to cut off because of time. They're giving me the, sorry about that. I'm gonna tell John he can't answer his question. So I- Oh, come on, John. You should ask a question. Because I want people to get out to the next session. So can you all stay for a few minutes if people come up and have questions? So all right, thank you for your time. For the panelists for joining us today, for all of you for participating. Again, if you have some questions, come on up. We'd be happy to talk with you. Thank you.