 But welcome to this panel you know I am really honored you know to be able to moderate this August panel in November okay you know a lot of different perspectives you know it's gonna be a lot of fun okay I just just before I get started I want to get an idea of you know kind of the background I think one of the questions that I wanted to ask was already asked right but the next question I wanted to ask was how many of you consider yourself as kind of application developers enterprise developers you know how many if you are developers per se have written I don't know thousands of lines of code of Java Python whatever their favorite language may be okay how many if you have never done in development and will never do a development you know will do a development across your dead body is something like that right you know how many of you are okay nobody okay that's good so what what we are here to talk about is you know should you know I purposely wanted to keep this panel a little bit controversial and in fact today morning when we had you know kind of a titty-titty to talk about the panel you know we said there are really no rules okay so if they fight you know it's just because they really disagree on it and they're passionate about OpenStack succeeding okay the only thing I said was if you can let's try to avoid some bloodshed okay but really in the end you know it's your panel hopefully you'll ask some thought provoking questions and you know keep the panel interesting I am the guy who adds least value so I'm gonna keep the introduction to you know as minimal as possible I know this is really early days of OpenStack you know we just celebrated the third birthday right and and really if you think about how the app economy the API economy and how contributed you know to the society in general if you will I think developers are are kind of in this unique position where they can you know influence the growth of pretty much any system right or the growth of any ecosystem and when I mean developers you know I need to draw the distinction it's not the OpenStack developers who you know whose life and blood is OpenStack right this is developers who are building on top of the OpenStack one of the things when I attended the Portland Summit was I thought that you know it could be either as part of the evolution but you know or for whatever reason I didn't find OpenStack as compelling from a purely developer perspective from a purely application developer perspective okay and like I said it's it could be part of the evolution you know there are some really cool things happening in Havana that obviously is gonna sway but but I think you know I think we have turned the corner I think application developers you know are gonna influence OpenStack and if you have to grow from hundreds of thousands to millions I think we have to have the application developer in there okay one of the things that I strongly you know and I strongly talk about this as a seminal paper in 1985 by Michael Michael Stonebreaker how many of you have read this paper by Michael Stonebreaker you know this was written in 85 where he talked about share nothing or shod everything okay and in fact OpenStack design tenants how many of you knew that there was an OpenStack design tenants you know yeah okay cool if you if you go read it you know it's pretty cool and I think the reason why I bring that up is as an application developer OpenStack is definitely something that's of interest you know if you want to do high availability you want to do scalability if you want to do performance which is really not the traditional strength of a core developer you know you would rather the infrastructure do it for you right so I think I think it makes a lot of sense to use OpenStack in this context okay and having said that I think I've already exceeded my time here okay so my name is Raghavan Srinivas I go by Rags I work for Ragspace I work as a solutions architect and one of my question that keeps me up in the night is should developers really care about infrastructure as a service okay and I will move on to the next speaker who is going to introduce himself Chris Ferris from IBM a couple of minutes hi I'm Chris Ferris I'm an IBM Distinguished Engineer and CTO for cloud interoperability which means that I have overall technical responsibility for all the open source and open standards work we do relative to cloud including OpenStack, Cloud Foundry and a bunch of other stuff Oasis Tosca and so forth my my primary interest is around interoperability that's that's why I took the title that I did you know a lot of what we do is really about it whether it's in the context of open standards or open source is really about interoperability and portability so that's sort of my that's that's my position and I'm sticking with it all right sounds good moving one over that's Diane Mueller and she's from Red Hat and Diane you got two minutes all right well thank you so I'm Diane Mueller I am with the OpenShift team which is a platform as a service that's sponsored by Red Hat and it's an open source project and I'm the community manager for that and a cloud evangelist and some people would call me a bit of a pause queen so I will put a few words in for a platform as a service here but one of the the burning things for me I think as my Twitter handle will show I'm a Python developer and an application developer at heart and for me what's happened with the rise of infrastructure as a service and especially with the OpenStack world has been the the rise of the developers involvement in the computing resources and the helping to build the requirements for those computing resources and having a say in how the cloud that we get to build and that we get to deploy our applications is built and I think one of the amazing things that's happening here at the OpenStack Summit that I'm very happy to be part of is a lot of the cross collaboration across infrastructure as a service the platform is a service you'll see I'm sure you'll be talking about some of the stuff that's happening at Rackspace with a solemn and with Docker but one of the things that's really key I think to where we're going as a infrastructure as a service offering that is OpenStack is that we're being able to take the developers use cases whether they're just web framework based ones or enterprise application developers who are trying to deploy very complex applications and make those use cases work within the framework of that OpenStack cloud and what we're seeing I think now is we've we saw it initially the rise of the public cloud with AWS and Azure and Google App Engine is being something that application developers like myself jumped on board with but now we're seeing people take that technology that elastic compute technology and bring it into the enterprise so I think now what you're getting is the enterprise use cases are starting to rise up for infrastructure as a service and that's what's making things really happen across communities. Okay moving on you have Lou Tucker from Cisco. Yeah hi everyone. Two minutes. Okay I'm Lou Tucker CTO cloud computing at Cisco and I guess one of the comments that you'd like to make is that we sometimes people pose us as IAS versus PAS and I think none of us on the panel would probably think that that's an or it's an and and that's because when you're developing new applications you actually are building on top of services and I also think sometimes we got it wrong it's really services as a platform that's what Amazon really perfected I think in terms of of EC2 and S3 and SQS it made the developers life so much easier they didn't have to worry about how are they going to store enormous amounts of data when they've got S3 or in OpenStack we've got Swift and Seth so I think that cloud compute cloud platforms right now have absolutely proven it's the fastest way to develop and to deploy an application you don't have to call up your IT guy to try to you know rack and stack a bunch of servers for something new that you want to do if you're running it even inside of a private cloud and application developer in a private cloud it's so much easier and that's why we're seeing the hundreds of thousands of applications that are running on top of AWS today and then we're seeing that now happening as OpenStack is now moving into a wide variety of markets I mean I'm seeing it going into financial services going into telcos going into so it's not just for the cloud providers but everybody's recognizing in the data center there's this new layer it's called a cloud platform layer and that's what application developers are targeting today all right last but not the least we come back to Adrian Anato from Rackspace and you know he's going to talk about what he likes about OpenStack. Hi I'm Adrian I'm a principal architect at Rackspace and I'm currently leading a software development effort called Solom in conjunction with about nine or ten other companies including Red Hat, Canonical, Ubuntu, I'll miss them all, eBay, Cumulogic, CloudSoft what we're trying to do is make the cloud easier to use for developers and specifically make OpenStack clouds easier to use for developers and to make applications more portable between clouds and what we recognize is when people talk about platform as a service that the systems that exist today for platform as a service are overlapping more and more with what OpenStack is trying to do and if you look at what's happened recently with respect to technology we have better operating system container technology than we did before that's really gotten to a point now where it's stable and and starting to become more prolific we've got more and more enterprises trying to do agile development and we've got this insatiable desire for cloud technology and so we have this gap in the OpenStack ecosystem where we want to fill how do we make the cloud easy for developers to make new cloud applications not just to take existing applications and run them on the cloud and so that's where we're focusing. All right I have a last slide and that you know this is it with respect to slides this is really your panel okay so whatever questions you ask is fair game okay you know make the panelists squirm in the seats a little bit okay so ask them the tough questions analyze what you hear from them and and just act on it so hopefully as a community we can we can contribute back you know it's not always that you have to contribute code you know to contribute to the community you can do something you attend the sessions in the track apps on OpenStack you know it's kind of like orange color something right you know this happens to be the apps on OpenStack track but that doesn't mean that there are a number of other sessions interspersed there I know Lou is talking on you know about networks and how they make a impact on application that's tomorrow at 12.05 OpenStack community welcomes developer in all languages every toes you know he's gonna be talking about J Clouds and a bunch of other things on Friday we have putting the pass in OpenStack in heat right and and there are a whole bunch of other sessions that I strongly recommend you go you know one of the things that Rackspace firmly believes are in core core develop I mean one of the core values you know is that it's full disclosure okay so in full disclosure I have to say that you know some of us have worked at Sun before right Lou worked at Sun Chris worked at Sun and I myself worked at Sun okay we know how that movie kind of ended so it's not what we are talking about right we are talking about different thing but there are a lot of panels that you can draw from that to this and and and hopefully you know you make this panel interesting so who wants to start off with questions questions yes good can we get a mic please please hold off if you if you can hold on hold on hold on second I think they may be recording yeah so just hold on one second if you can be patient maybe I can use this mic if you don't mind do we have an extra mic you know while we're waiting it just you know when you mentioned Sun I remember one of the things that we learned there as part of the early Java days was that when you create a new platform layer it's classic computer science you create innovation above and below the platform and so that's why we're seeing heat and everything else I think start to arise above sort of the core platform and we're seeing software to find networking and programmable infrastructure and everything else happening below it so I think that's something one of the parallels I think from the past to learn from question actually I just want to kind of qualify a little bit actually myself the developer and I think there's IAAS and PAAAS to me IAAS doesn't really it's irrelevant to me but the PAAAS is starting it does because that platform which I need to worry about so I want to know what your opinion on that and because I think you guys asked earlier right seems like a developer maybe some of them are irrelevant so I want to see you know where I am okay I have my own opinions but I'm not going to talk about that yeah let's start with Diane ladies first so so one of the things that I always say about someone has a cloud initiative and they if they only focus on the IAAS level then when they get to the end of that project they've even they've built a private cloud in in-house and they've they put their infrastructure and they have all their compute resources and then they go to talk to the developers the developers like you myself included are looking for the platform as a service and it's not there and so I think what we're seeing now is in in my humble opinion probably not my employers completely yet but is the convergence of platform as a service and infrastructure as a service so in order to really make infrastructure as a service useful you need things like heat and the orchestration tools and the platform as a service brokering functionality so that you don't have to build and orchestrate your own lamp stack and scale it not to scale it and that's really what the PAAAS functionality is and some of the technology that's coming out with heat and Docker and solemn if I say it right solemn is really about the next generation of platform as a service and that's what's being built and talked about here at the OpenStack summit this time around is we're getting the new building blocks for building PAAAS functionality on top of OpenStack and that's what makes OpenStack really useful to developers. Okay Chris you want to go or anybody wants to add or just want to make sure we don't over rotate on this because I think it's one of the tools developers use our tools and what we are getting with the new PAAAS methodologies and everything else actually is a better way to describe complex applications and their interaction so I very much like that but it's not you still have sort of the needs for a very large scale key value store you need have the need for something that which will dynamically provision virtual machines or Linux containers for you so these things are all become I think we are seeing that merger it's all the services these if we start thinking what we've really made the shift to is to services we're no longer talking about you know exactly you know a physical host that you have to now make sure the bios are set correctly and wire up correctly you have an abstraction that you can deal with which is and you're contacting a provisioning service a compute service and we shouldn't forget the fact there are hundreds of thousands of applications that have been built on IAS look at Amazon today Netflix built their own orchestration layers themselves so I think that we're you know the constant evolution of tooling and everything else is great and that's what we really need to see but I think in fact what you'll find you'll use a variety of tools and you'll use the tools that are most appropriate for you and if you fit into actually you know you remember like in Heroku I was using Heroku for well me that that sort of model that's great I loved Ruby on Rails that was another environment of framework that I could rapidly deploy an application but I then was the constrained when I needed to actually scale out to a certain dimension so these are tools and we should use them all and really understand the capabilities of each I And But they still have a need Whatever it needs and I'm the line substrate services that deliver Is that enough for you or do you need something along the lines of Or do you need something that Princess Oasis Tosca can give you to Essentially virtualize an existing workload Sometimes a complex topology of services that are all interlocked together in frankly, you're not going to rewrite that anytime soon, but you do want to virtualize it you want to be able to put that up in close to an inner cloud but you want to orchestrate how that thing is started up because it's a delicate process and and so you need something way young Whether it's open shift But that's again you're not that's not for an application developer That's something that's necessary and again That's where we have orchestration tools and so forth that allow us to stand up those complex workloads So you need this range of capabilities So I don't think that there's ever gonna be one answers like you said Lou It's gonna be there's gonna be a collection of of tools that we can bring to bear, you know, whether it's Netflix OSS or whatever To deliver the value that we're looking for in our particular use of OpenStack So I'm gonna answer the question with a question How many in the audience are on a dev ops team where you operate and develop the software can you raise your hand for a minute? Okay, this is still a Growing trend so I'd say maybe I got ten hands As You see more and more dev ops This question starts to become more and more relevant. Okay, when you have peer application developers They really only care about having a place to run their application and not worrying about infrastructure Those are kind of their key interests, but once you start going down the dev ops track You start to care about the infrastructure more and more and you start to care about automating things You start to care about things like test gates. You start to care about continuous integration continuous deployment and If you care about that stuff Then you don't really want a separation of concerns between platforms and infrastructure. You would like something more blurred and More general purpose so that you can fit exactly what you need is a dev ops engineer to get what you need from it So I'd say where we are today. You're right. You probably might just want to focus on pass But where we're headed and where I believe we will be in a matter of years Is that a point where we don't really draw a distinction between the two? They're really one in the same All right now that we have beaten the is versus Let's go to over there So you're all talking about tools for developers some blur some of the most fundamental tools We can provide to developers our software development tool kits, you know these things written in specific programming languages that talk to the OpenStack API and Then others can build more complex tools on top of those but you really need that that fundamental layer there So my question is how do we get OpenStack contributing companies? Contributing to these open source SDKs out in the wild like for example fog chef is Dependent on fog puppet is dependent on fog for spinning up resources on OpenStack clouds How do we get the Cisco's the IBM's the red hats contributing to the wider tool ecosystem in OpenStack? We we do There's the SDK question and Who wants to take a stab? I'll answer it is from Rackspace perspective Yeah, Rackspace contributes to fog. We also have our own SDK that suits our own our own narrow interests But they absolutely do and the reason why you don't see huge amounts of contribution to those things is because they don't They're done. They're they're finished. They're not There's just not a lot of action happening in fog right now You create new resource types right and you get new new abstractions, but fog is a finished product. I Would disagree. I mean there's constantly new API is coming up in OpenStack So you constantly need to and there's changes as well. So you need to add and maintain that coded fog. That's happening, right? All right moving down right ahead One thing I would just sort of say OpenStack is not going to cover all of Software it should not We have to you know, there's a lot of other very You know vibrant open-source projects that we're looking to make sure work well with OpenStack and are available And we want to also make sure there's commercial software opportunities for the companies that doing it I think your question actually gets to a deeper point on me if you You know IDEs, you know something that is really integrated on your desktop deployment on top of a cloud And that's where I see a lot of the past work integrating back into the developers You know desktop so that they can with containers and things like that We can start to do a lot more of these things So I think you're going to see it but it won't necessarily be an OpenStack project What we want to make sure is that these things all work with OpenStack because otherwise OpenStack just grows sort of uncontrollably And we have we really want to keep the focus on OpenStack on doing the things that it's really set out to do better and better and better And leaving opportunity and working with other communities to get a lot of the tool tool sets around it as well I still see that look on your face. So yeah, I'm just I'm just a bit concerned and in the OpenStack community of kind of the obsession around Commits to open github.com slash OpenStack when yes, all that all that super important But OpenStack to a point where you know, we need to provide that fundamental just as AWS provides their SDKs Same thing we contribute to chef we contribute to the job SDK, I mean There are a number of things that IBM is doing that maybe is not as visible We don't you know Talk about in the context of what we're doing in OpenStack and yes in the broader in the broader context I mean, we're all about I open stack. We're you know, this is our strategy, right? From an I ask perspective. So whatever we need to do To enable it for our customers. We're we're contributing to yes So I would I would just say like at Red Hat. There's over a hundred thousand open-source projects that we participate in in some way and There are ID's that are integrated already with OpenShift and you know as far as that matters Then then that means it's running you can deploy to OpenStack or AWS or wherever you want from it So it would when when we talk about Participation in an open-source projects. I think that we're so spread over so many different projects that I don't know exactly who's contributing to fog from Red Hat, but they may even be in the room or here at the conference Okay, thank you. Thank you Moving on next question Yeah So one question I might like to get back to it. Okay, Robin Yeah, you guys talk about from an applications developers perspective being able to more or less provide a structure to where we can use any Provider and be able to transport from provider to provider. How are you? How do you go about? Ensuring that when I write some code that I can transport it and that it's that it's mobile Is there any way is there like a tenant that's out there that says that thou shalt follow this interface so that? It makes it easy for me to transport my code Okay, well the there is a definition around open so there are open-stack API's And we're having in there's certification and we're looking at you know applying the power of the brand To ensure that we have interoperability between open different open-stack deployments, whether it's on a private cloud or at a service provider and that that's a lofty goal though because You know that takes time for everybody to sort of come up to the same release and everything else like that And so perhaps you know on the on the inspection side on a tool You might be able to develop some of the technology to find out whether you're at good deploy but I think one of the one of the real tenants of open-stack is we see The real that one of the real value propositions of open-stack is that we hope to see hundreds and thousands to thousands of open-stack clouds out there That are both you know in-house, you know a private cloud and at service providers And we want them all to have the ability the developers to move their applications freely from one to another provider They have the right credentials and everything else Right, that's exactly right. I mean again We're looking for that interoperability and portability Across all these open-stack clouds and that I think you know applies equally to the tooling that gets applied in the context of open-stack and again whether that tooling is developed as a project in open-stack or a Tendent to open-stack if you will like you know some of the stack-forged stuff or you know if it's happening whether it's an open shift or cloud foundry or Whatever right Or or Eclipse for instance if you have just a an IDE You know we're working on you know things in jazz up and the intention would be that you can interact directly with the underlying Cloud if you will I mean shouldn't shouldn't really matter was an as or a pass So some of the work that's being done on Docker and you'll see if you read up on that the Containerization approach that is now being adopted into open-stack as a project That's going to help a lot with the portability of your application across different open-stack clouds and and other clouds that Adopt and support the Docker as I hate to say the word standard yet, but but I think but I think what it's an emerging Area and I think if you keep an eye on what's going on with Docker You'll see that aiding along with the certification and keeping the brand But open-stack itself can be deployed in so many different ways That I think and Randy bias who was on just before this really put it put it well and saying how Open-stack is it's not quite yet a system It is a framework of components that people build clouds with and trying to keep that to be a Standard way that everybody could build on a private cloud a public cloud or a hybrid cloud Applications that would port anywhere is a lofty goal still Is this working? What is that typical size of an open-stack deployment that you see let's say the the minimal one that you see 10 servers Or something like that or one even Yeah But seriously, I think the question was what is the typical, you know I mean definitely the minimal is one but the typical if you can describe the various deployments that you saw that would be helpful Usually it's a couple racks happened yet We did a survey of users and there's some of the data in that which show going up to you know the largest there are You know in tens of thousands A small number of those actually interested in a smaller Custom well it we I've got an open-stack team, so we you know each developers sort of gets three blades So you can you can do quite a lot with very small, but I think that the interesting aspect there is actually memory on these systems is growing quite a bit and so on Blades or rack mountable where you bet a lot of memory you can run, you know hundreds of virtual machines So you can do quite a bit with just a small, you know half rack cloud or whatever it can be pretty substantial for a company just starting out The number of noses usually proportional to the amount of storage in the network In a typical cloud, so if you've got a tremendous storage requirement, you're gonna have a lot of nodes It's gonna be a lot of gear If you're using it mainly for virtualization, then you'll have fewer nodes So you typically see one room one room or less full of equipment If you've got a huge storage requirement and you're trying to store a petabyte of data, you're gonna have you know thousands of machines Okay, I do have another another question. I'm not sure it's the right audience. Do you see windows relevant at all meaning? Sure About half of what racks based cells of windows Okay Next question Yeah, thank you if we say that an enterprise app developer is building like an app That's gonna scale up really large and it's gonna have lots of different types of nodes What can we do to help people to understand the interactions between the decisions? They make on the compute and the storage and the networking that as they kind of try and get that all to work together across a large area Yeah, that's a great question I mean, I think if you if I understand you're right, you know, how do we make those decisions because open stack doesn't recommend? You know this many storage nodes this many compute nodes or whatever, right? You know one thing that cloud computing does not eliminate is a need for good design You do have to think about this. You have to incrementally roll out your functionality test Observe the behavior of your system and go forward there There's nothing's changed in that regard that I've seen in the last 15 20 years So I think what you're getting to though what we could all benefit from as a community I think is much more sharing of examples and with real performance data real scaling characteristics I've constantly on my team asking them to go and try and make those measurements right Shannon and Publish them the result and I urge everybody in the community because I think we'd all benefit from them More reference architectures, you know things like that. Yeah, so I want real I want real applications Right real performance and how they designed it and everything else my real benchmarking is There is that but I and again it sort of it gets to the to the point that there's sort of Maybe this is a little bit too black and white But there are sort of the class of workloads that you know It's the 20% of applications that matter to the business frankly really, you know They are the ones that actually bring the money in and those are the ones that the it really needs to care about And they want to babysit those they want to optimize optimize those systems You know to be as performant as they possibly can be Because they need to be and they need to make sure that the damn things don't fall down, right? Absolutely positively and then there's the 80% of the applications that they run that frankly Are an annoyance to them, right? They aren't the ones that bring the money, but they're necessary, right? And then but do you have to invest all your resources to optimizing those things, right? Or is good enough good enough for most of them? And I think you know again This is where there's a lot of experimentation going on that you know you take a look at something like Cloud Foundry there's one way I mean the developer does not get involved in the architecture for how storage talks to networking They don't know and they don't care and they don't want to know frankly, you know, they just want to write their damn app and I need Databases of service. I need you know, I need to be able to sort of suck stuff out on Twitter You know, I want to have you know an analytics service, whatever it is That's all they want to deal with they don't want to deal with what was the architecture of the storage solution that got My data, you know as quick as it possibly could to my application because it's not that important to them Right and so, you know, I think you want to have this range of capabilities And so there are going to be that there's going to be a need to tune your ERP system or whatever To you know to a fairly well and then there's going to be a lot of the sort of the more systems of engagement style Applications the mobile social and all that kind of stuff that frankly the developers don't want to know and they don't care And and and we will get these things to scale horizontally and and those other problems become less of an issue for for the application developers All right actually just Horizontally is a hard problem and Fortunately what we have is that we're built on top of services that like a key value store or Message bus system or whatever which has been designed by people who really understand what it means to scale So the individual developer doesn't have to and that's another attribute I think it has and everything else is that people are designing those things who are thinking a lot about the architecture and about The scaling properties of it so the developer doesn't have to they can care about their application Really the only thing the application developer Cares about at the end is the cost of those compute resources for their application and so Managing that and having that work on the auto scaling of their application and making sure that they're getting the best bang for their buck is really what they they care about the most as long as That that piece of the puzzle is is taken care of by the platform as a service and their application scale up and down Successfully and they don't get charged for resources and not using they're usually happy Yeah, I'm not sure if I agree with that because when I was an application developer You know, I just used to spin off VMs and of course my manager would come back and worry about it So, you know, but in any case, yeah, let's go back over there Somewhat of a different topic one distinguishing characteristic of some enterprise applications is regulatory compliance How do you see that playing out in open stack today? And where do you see that going in the future? Excellent question regulatory compliance because you know like it or not That's the problem the lot of enterprises face, you know, he power, you know, all those different standards, you know How do we deal with it? It's a real question. I mean, I think that two things have to happen One is that regulatory compliance is have to be adapted for cloud computing Because the idea of actually going in and doing an audit and things like that on Amazon doesn't work So we have to understand the intent of what we're trying to do with with the compliance and then I also think Even we take virtualized networking that was a lot of what, you know, if you heard Martin's talk I mean he's talked about you're recreating it so that you can be compliant so that if you demand isolation of this Vlan well a Vlan was that that's what it is a virtual land. You're using shared wires, but compliant regulatory You know people have understood that that's okay There's isolation there and that now when we're saying now we're gonna be using overlay tunnels instead and everything else We have to be able to prove that there is that same kind of isolation so it can pass those regulatory compliance regular, you know requirements Yeah, so there's there's certainly the Can this be certified for these ma have a fed ramp whatever right or and I'm using yes But obviously there's other regimes that apply but typically those those regulations apply to a Running solution so it's not the software necessarily that you can certify It's the actual environment into which you poured it right and it's the operation of that particular environment that are change management and all of the auditing everything else so So so very much to lose point Some of the regulations that we have on the books today, you know, wherever they're coming from are Frankly, they're somewhat dated and they need to be sort of rethought again to get at the original intent right and to help The you know the government officials that are coming up with these things to better understand What this new world really means and and get at the core of what the you know, the various requirements are today So that's a lot of what was some of the think that was going to Fed ramp originally It's not done, you know by any stretch of the imagination It's still got a ways to go but it is being but it's better than and it is being taken up and it's very serious I know I was on a advisory panel with the Vekundram when he was saying there's cloud first You know policy in the White House and that you know government agencies had to Justify why they weren't going to take an application to a cloud and build a new data center they were trying to stop the proliferation of a million data centers and All of us in the industry could say but regularly compliance You have to you have to address these things and you have to have those agencies update their thinking to accommodate that And he said by having a cloud first policy. That's wedge. We're gonna use So I think that we're gonna see it, but we know those things take a long time a very long time constant But I think I think we'll get there, you know And it's not just the sort of the the governmental regulations and so forth that apply It's also the corporate regulations. I know I'm running into this being IBM Probably not a surprise anybody There's a lot of process and you know, you have to go back to again You have to go back to the original intent. Why is that process there? What are we checking from a security perspective and then let's actually work back walk that back and understand what it is that we're doing in provisioning something based on cloud or How we apply a context like DevOps When you know, it's typically you have to isolate who you know operations from the developers, right and the developers They can't get in there. No, no, no, you you can't do that It says right here You can't you know and so we have to change a lot of that right in order to get to where I think a lot Continuous deployment and you're deploying maybe 10 or 100 times a week and it takes two weeks to actually approve a change Right, that's not gonna work. So those things have to change But look at the intent and I think that we can still achieve the same kind of Compliance that we want to with both the government and also corporate regulation All right, I think we have time for one more question, you know rather than end on the topic of compliance You know, maybe we should end on something else, you know, hopefully One question one more question going for ten dollars Does anybody have a great great story about how their lives were made better by building on top of a cloud instead of in the old way No, you asked yours. Let me see if anybody else anybody else questions No He's gonna get the mic again. Watch out. Yeah, all right. Yeah, I'm just if you can Open stock versus a recent let's say do you see Customers replacing it's an addition to what they have. How do they manage both environments together? Yeah, are you are you asking this question from a developer perspective or more for me? I'm a I'm a software vendor I'm a vendor so perspective vendor. So how do I foster an ecosystem something like that? What do you see customers doing? Okay, so a lot of customers already have these center ESX deployed. Okay, you have open stock What are they doing new applications new deployments new Server rooms are they working? Together or just separate applications any experience there? I'm asking any question for it from who wants to take that I'll give a very short answer I think most of the new applications are going to a cloud platform and I think the migration of quote Legacy applications into a virtualized environment or a target of the v-center So I think we're going to see both existing very for quite a while Because I think they serve two very different different application sets Yeah, and then I think that there's certain application of open stack and some of the cases we've seen where They already have v-center and they want to be able to sort of start Nibbling away at that environment in the context of open stack Have open stack do the management right as opposed to v-center And so they want that integration and whether or not they then branch out from there and I also adopt KVM or Zen or whatever You know different story, but you know they start by saying I've already got this I want to take advantage of that All right, and I want to be able to use I want to be able to use open stack to manage it Okay, do you guys want to add anything or? Okay, how about we do a closing remark or are we done? What do you think? Okay, I see that it's done. All right So I really want to thank all of you for you know for asking the thought-provoking questions I really want to thank the panel, you know, I know they had a very busy day very busy week And they are here. You know, just seek them out. I'll put up their Twitter IDs as well You know, it's right in the first slide So feel free to hit them up on Twitter. Okay, or if you find them anywhere, you know, just talk to them Okay, enjoy the summit and again. Thank you all for coming and thank the panel again. Thank you