 Okay, thank you everyone for coming to this session today. My name is Lou Tucker. I'm a VP and CTO of Cloud Computing at Cisco Systems. And I thought today would be interesting to look at how one company, Cisco Systems, sort of adopted OpenStack how we got involved. And so maybe you can draw from this some of the experiences that we've had that will allow you to bring OpenStack into your own company. And so I want to be able to also show off some of the technologies that we're using and where OpenStack is being taken by the different business units within Cisco and hopefully leaving time for at least a question or two at the end. So we very much believe that OpenStack is where the standard is being defined. As you can imagine, Cisco has been very involved with a lot of standards bodies over the years. And now we see open source really coming into the fore as being a place where we can develop these through code and then share them with the rest of the community. And so we participate in the very first OpenStack Design Summit in Austin. You've seen the graphs of how those things have grown every time. It's great to see 6,000 people here. It's great to actually see a number of faces here that I recognize from previous summits. So I think that we all should be very proud of what we've been doing. And so this is when we launched and got involved with the foundation itself. I also serve as vice chairman of the foundation. And that's been a rewarding experience because as we see all of the different companies, vendors and customers and users getting together to define how we are going to evolve this platform and solving the issues that we see present in the marketplace. You know, to win anything, you've got to play. I mean, you've got to make a bet. And at the time in 2010 when we were looking at this, this was not an obvious bet. And for a company like Cisco, it was certainly not obvious why Cisco should be involved in an open source community such as this. But we made the bet. And what you can see is that then you can talk a lot about how do you affect change within an organization. You have to be able to start with that. We started with a very, very small team. We set ourselves up to actually have a very large number of friends and family, people in the rest of Cisco. Cisco's got 70,000 people in it. We were a very small team, half a dozen people. And we had many, many other engagements with other participants from the rest of Cisco getting involved. And that friends and family, I think, was a really important aspect of getting that change. We also had to understand the differences between open source and proprietary software. And that's where I think that we're seeing real resurgence of interest in open source. I was just actually at Onog in New York debating open source versus closed source. And I felt kind of sorry for actually Charlie Cinkala who was taking the proprietary side because he had to defend proprietary software. But the best we could come up with was that when you have the developer community like this, open source is the preferred model. This is how we can really get the innovations and get the brightest people in the planet working together on something like this. That, again, we are all in businesses. And so you have to be able to show the relevance to the business, to do this, and then be able to rapidly track technology trends. This is moving faster than anything else I know today. And so you have to be able to be very agile and move quickly as the technology changes and as the marketplace changes. So in fact, if you look at where we've been making contributions, we're primarily known for contributions in bringing in neutron networking into OpenStack. But again, it's larger than that. It's being able to expand into a lot of different areas here where you can see now we're focusing a lot on containers and making containers a viable alternative to virtual machines running in an OpenStack environment. IPv6 is very important. And so we're making a lot of contributions, core contributions there in terms of how IPv6 would be made available to everybody. In addition, then there's other things in terms of Kafka and how big data pipelines are being able to be used. And we can see a lot of transitions, different parts of the industry are moving to OpenStack. And so the use cases are exploding. And we're finding the right projects to be able to bring forward that kind of technology. So it's always about innovation. And here's just a couple of innovations that we've come out of my team at Cisco. One of the interesting things, these actually, again, started as very small projects, often by interns, that we allowed them to go off and do something, such as curvature and what we're doing with analytics. And we also noticed that there was a real lack of performance benchmarking. How many of you really know the performance of your cloud? We didn't have many tools which would allow you to measure performance. So we've been contributing those. And in each case, we've been working on this technology and developing it in the open, making it a broader than just what our own use case was going to be because we can get the rest of the community involved in it. So for example, in the predictive analytics and visualization, these are large scale systems we're dealing with. And the traditional management tools just don't cut it. You have to have alternative ways of visualizing the information and visualizing the data. So this is where I find it particularly intriguing because a lot of new developers who are not very deep into infrastructure know an awful lot about how to do visualization. They know how to build web apps and web services. And that can be readily built on top of this. Recently, we also started to discover as we're putting more and more projects on top of OpenStack, we're putting, as we will see in this talk here, some of our video assets and mobility assets, what they want to know first and foremost, does the cloud work? Before they're launching their application, are the services there that they need and they can rely on? And what are the performance guarantees they can realize on top of OpenStack? So we came up with something called Cloud Pulse. And I think of this as this is a health check. This is something that would allow us to look constantly at the cloud. All of the services, a lot of different metrics in an open and an extensible way so you could add in your own monitors. And this, I want to actually think about this as having sound because if you can just sit there and you're hearing a beep, the cloud's still alive. That's good. So I think there's a lot of these things that we're seeing coming into the forefront again. For example, with Ceph, there's an awful lot of information in Ceph about how the data is distributed across the different Ceph nodes. You want to be able to look at that. And you want to be able to know when those different critical points are being met. So dashboards are becoming very important into that realm. Excuse me. So of course, the stats is what everybody looks at. And so when you look at number of blueprints, number of contributors in terms of lines of code, those are all important. But the most important thing I want to stress is that this represents a total commitment. Is that it's not just putting in your own code. It's helping other developers with theirs. So the first advice I give anybody who's new to OpenStack is before you submit your first line of code, review five others. If you start that process of reviewing other people's code, the whole community will move faster. And guess what? You will then have people willing to review your code. We are very much of a social organization. And it starts with this review. So I look, most importantly, how many reviews? And we have some of our reviewers in the top three or four of the reviewers in the community. Excuse me. So how do you go about driving change within an organization? First, you have to really understand what you're trying to do. And one of the first things we did was to get an application. That first application was WebEx. Many of you know that application. And you saw that that was one of the first user stories that we had up on in the OpenStack community. The other thing is you have to learn from your customers. So Comcast was an early customer of ours. And instead of saying, here's OpenStack and throwing it over the wall to them, we worked with them. We said, we'll help you become part of the community. So that not only are we giving them technology and assistance, but then they were becoming active contributors. And as a result of that, you're seeing that Mark Bule up on the stage today. You're seeing that Andrew Mitry and his team is actually launching X1 and XVINITY on top of OpenStack. And they're also some of the core contributors into the IPv6 arena. So this is how you really want to work with your customers. This is the co-development that we see with our customers. And that's the way Cisco likes to work. And I think that it's successful for you folks as well. Sorry, I should have built this out. Excuse me. The other thing is that you do have to work with your field and the different business units. Do not underestimate the amount of education you have to do within your company. They are going to be approaching you at all different levels. Some people will know an awful lot and want to get very involved. Others will think, oh, OpenStack, you install it just like you would an app on your iPhone. So there's an awful lot of effort that you have to pay attention to. And to try to work with each of the different business units that you have so that they really understand what OpenStack is and set their expectations accordingly. What you also should be doing in that is recruiting other developers. This is about a community-developed software. We need more people developing software. We need more people contributing to this so that always as you're engaging others within your teams, use that mantra to get them involved so that they can actually be self-sufficient. And it also takes a village. And so one of the things we've been trying to do over the last four years is building an internal community within Cisco. So it's not just my own team, but also we've been having a summit that we hold within Cisco where we actually share our best practices. It also allows people who have been working in very distant parts of the company on OpenStack to get recognized for that. Because we did it just like everything else. We've got sessions, we had papers contributed, we had reviews of that, and then we posted everything back internally online for the long tail of the content that was produced from that. And we're going to do that every six months. I expect it to get bigger every six months that we do that. We also have, it's important, people have talked about change to get your executives supporting this. So we went to our executives, and we filmed them talking about OpenStack. And so we end up with this. This is a large now conversation taking place within Cisco by the leading executives of the company. This goes a long, long way when you're trying to convince other parts of the company to work with you and adopt OpenStack. So we did have our first internal summit at Cisco. And we had over 400 people attend over 30 different business units. I was personally surprised. I didn't know about all of the uses of OpenStack within Cisco. And it's growing every day. So this was a great way for us to be able to model it after this larger community summit that we have internally within the company. And this will be then an ongoing event that we have every six months. We also have our own developer conference, Cisco Live. And this is the number of tracks that we have in sessions that we're talking to Cisco's developer community. And this, I think, is a very important way because each one of us have either our own communities or our own industries that we're working in. And this is a way to actually bring OpenStack out into those communities. So when we look across OpenStack at Cisco today, we can see it's not one is that we have it being used as a cloud within our IT services and things like that. We also have made big announcements about InterCloud. This is a global network of clouds with our service provider partners that we are federating into an InterCloud framework so that now through that kind of alliance, businesses we're able to get services in geographically distributed areas around the world on a common OpenStack platform. We also have big pushes right now into NFV. Obviously, that's a big interest to Cisco and a lot of our partners, something called Cloud VPN, which is being able to provide virtualized network services on top of an OpenStack cloud. WebEx became Spark. And the Spark application now is running fully on top of OpenStack today. What I wanted to do in the remaining time, though, is to go on and talk about one of those use cases, which is in the video space. So if I can now ask Gaurav Rishi to come up, I think we can go over how OpenStack is being used to support our video. All right. Thank you very much. Thanks, Lou. So like Lou said, there was a small bet made, which actually has grown over time. And one of the bets was actually in the context of media and entertainment. So in the spirit of sort of learnings and teachings, I think what I really wanted to do was, in the brief time I have, talk about sort of the why as a business unit, we decided to sort of look at OpenStack and kind of share a little bit of a journey. And then I'll also invite a couple of colleagues to come and actually show it to you. And we have live demo, if you will, in our booth. So this would be a good way to get and see it. So if you look at media and entertainment, so this is your service providers as well as your media companies. Now there are many things happening in that ecosystem. And I've kind of gone to different conferences and focused on different types. But today I think we just talk about, all right, what does cloud mean in this context? And how has Cisco and OpenStack do we sort of contribute and sort of take our customers forward on that? So if you look at it, if you take a step back, and this was sort of maybe to some extent an evolution, which is what is the problem today? I mean, what are you trying to solve from a business perspective? So if you look at the proverbial stack, if you will, today almost everybody thinks there are unicorns, as far as an application is concerned. There is a static sort of siloed application. Maybe it's an encoding application. Maybe it's an origin server. And they use a particular type of compute network or a storage resource. And the organizations, as Lou said, are also siloed even within our customers. So a lot of time when you want to go ahead and intellect these little functions together, actually it's a very manual process. You do see people sort of passing Excel files to and fro just to kind of create these linkages. So they're not truly dynamic. And out of that, ultimately, as you see and you see it in the keynote discussions today, there is a fast pace of innovation or a faster pace of innovation in terms of the experiences that you are as a consumer expecting. So whether it's multi-screen video, whether it's Timeshift TV, whether it's Cloud DVR, or whether it's live to what, I could sort of go on about that. But the point is the pace of innovation there is changing. And what ends up happening, which is good news for some of our services colleagues, is that especially when you have multi-vendors, it turns out to be a complex service integration project where you kind of have these silo pieces, like I said, and just bringing them together not only is slow, it's also pricey. So that was actually really the business problem which made us take a step back and say, well, we've been in this business for a long time, and we've probably been guilty of producing some of this stuff. How do we go ahead and actually fix it? So I think, again, it's not a magic bullet or a silver bullet where we kind of go ahead and sort of fix one thing. It's about going ahead and sort of looking at it holistically. So the three or four key points I'll just sort of quickly make is, one is I think we wanted a programmable infrastructure. I think OpenStack was very little debate, frankly, from our perspective. You kind of look at the ecosystem around you. It was very much given that, hey, look, we want something where we can actually go ahead and quickly, if needed, tweak the underlying infrastructure, optimize it for video, because if you really talk to the people who were doing video development for the last 15 years, they would say, look, we are special. I mean, we can't handle the jitter that is expected off these legacy qualm devices, and you still want us to go ahead and deliver video there. So we wanted the ability to be able to go and make the tweaks if we needed to, whether it's a storage part or whether it's the delivery aspects of it. The next step actually came up to be the actual application functions. And so Lou mentioned sort of NFVs. We like to be special, so we call them VVFs, which are just video virtual functions, but it's basically the same thing. And essentially what we said was, let's go ahead and rethink about how do we think about encoding in this context, and how do we make them modular enough and also pluggable enough so that we can actually now quickly software these applications so that you can use workflows or orchestration and heat templates to go ahead and actually very quickly get to the next step, which is actually the services that our customers or their customers in turn really care about, because that's really what generates revenue. And there were other fundamental sort of takeaways we had as we started getting into this journey, which was essentially, for security, for example, was something that was typically endpoint security. And we've invested a lot of money both organically and inorganically in that. But once you sort of move your ecosystem over to an inter-cloud infrastructure where you're moving workloads and dynamically provisioning origin servers or encoders, like I said, or packages, you need to be thinking about security from a completely different context. And that has to be vertically integrated. So that's something that you'll continue to see us sort of evolve on. And I think when you talk about it in the context of zooming in further, I think there's a lot of talk about here in terms of container. So I just want to sort of distinguish. Here we're talking about sort of logical applications, but I think you can join the dots very quickly to figure out how this is sort of going about. So when I talked about network function virtualization or VVFs, as I call them, each one of those functions that you talk about, when we think about evolving them, and this could be for a completely different, you know, I'm talking about media and entertainment, but I'm fairly sure it's applicable to a lot of other segments. Each one of these functions now, in our context, we said, look, let's look at the next wave of modern application frameworks. Think about these applications actually as functions which can sort of scale horizontally. They are stateless. You can go ahead and actually software them using these media workflow aspects of it. And by soft chaining them in different permutations and combinations, or maybe, the chain of one going and feeding another one ends up creating sort of live workflows. So if you have live, what are you doing? You're just taking content in, you're encoding it, probably applying some DRM, doing some packaging for different types of devices. But now when somebody says, well, I also want to solve Cloud DVR, you're doing a lot of that, but you're also adding this new step of going ahead and storing it and being able to record the content and then play it out and maybe reformat it in a different way. So you end up reusing a lot of these components, but what was broken in the previous model was organizations, first of all, were not ready, and then the product itself in terms of being able to take these functions and software wasn't. And so this gives you a construct to sort of get there. So there is a lot more to it, but what I wanted to do was maybe bring in my colleague, Saravanan, who's actually got some of this working. We invite you to sort of come and see this at the Cisco boot, but let me ask him to sort of give you a quick view in terms of the challenges that we faced and some of the resolutions. Be great. Also, explain what Cloud DVR is. I think that we talk about it, but I think that the audience thinks that there's DVR on their own. Can I have a quick show of how many are from media and entertainment as a, oh, this is fantastic, very few. So that's great. So let me talk a little more. So think about, I think DVR is something that probably 35% of the households have. And so when you talk about Cloud DVR, it's taking that same functionality, but moving your storage into the Cloud. Now, why video we think is still special is because if you think in terms of bandwidth or in terms of storage, the amount of volume is huge. I mean, as an industry, I think that's where a lot of the requirements get driven out of. So Cloud DVR is taking that functionality. The advantages, of course, are you get geode-independence. You could be watching this and sort of watching the seaplains land, your content that you've loaded. Also, you can go ahead and actually get to watch it on any kind of device. So that's a given. Now, if I take the next level of peak under, from a technology perspective, you typically break it up in terms of data plan and control plan. So data plan is the heavy functions where you actually touch the video content. You might be encoding it. There's a lot of bits moving around. You might be actually going ahead and repurposing it to kind of go ahead and insert advertisement, which is sort of very much local to you. And control plan are functions which are essentially where you don't really touch the data path per se, but you need it essentially to go ahead and make sure that you are the authorized person to be able to record a particular piece of content. Or when you schedule something, it goes and sets up a little timer inside to say, all right, I need to make X number of copies of this program for this user. So there is a logical separation. And I think when you think about from an implementation perspective, of course the requirements that you need from your underlying infrastructure and functions are very different. And finally, the third piece of getting this cloud DVR together is the client device. So ultimately, as a consumer, you're seeing this cool user experience. You want to be able to not get bogged down between scrolling through a linear guide and get recommendations, et cetera. And that's sort of the client device. And these are the three key components which actually end up making that solution. So with that said, hopefully that's sort of the quick primer on sort of just the solution. But with that said, let me sort of very quickly ask. Can we take questions at the end? Oh, all right, yeah. If you can wait for a few minutes. Thanks, Garo. Lou talked about all the different business units inside Cisco are working together now to actually bring in a lot of applications to run on top of OpenStack. One applications, obviously, we were talking about is the cloud DVR and other media applications that we're putting from a hardware-centric to a software-centric applications and also getting them ready to run on OpenStack. That's the key thing that we're actually focusing on as far as one use case is concerned, which is videoscape as such and all the media applications. So when we actually broaden all the applications, hardware-centric applications and virtualize them, that is the first step that we did. And the next step is virtualized applications have to be cloudified. Not all virtualized applications are cloud-ready. You have to make a lot of changes internally at the app level to make sure that they're ready to be run on cloud. In fact, if I might just interject here, a couple of weeks ago, there's National Association of Broadcasters meeting in Vegas. Big show, lots of big screens and flashing video around and everything. This is where all of the major video producers' content meet. That's where the industry meets. It's an industry that's been driven by hardware appliances. So everywhere you go, there's specialized pieces of hardware, whatever, for encoding, transcoding, ingest, and everything else. There's two kinds of vendors there now. One kind, about two years ago, they recognized something's going on here in cloud computing and they can start to virtualize those. They had an aha moment. There's the other companies that are sitting there with their hardware appliances and they're having an oh shit moment because they're going, this is coming and I'm not ready for it. So I think we're gonna see this. We've talked about disruption. Jonathan talked about it this morning. This is an example of a whole industry now being disrupted by cloud computing. So I'm just going to take one slide to explain to you all the challenges as a video application team be faced when we put all the video applications into OpenStack. And as the slide deck shows, we have a live working demo in our booth where you can come and experience the video applications running on OpenStack, live streaming, cloud DVR, WOD, and all the use cases, which used to be run on dedicated hardware. Now it's virtualized and it's cloudified and it's running on OpenStack. So just a quick challenges. Obviously multicast routing was a big challenge for us to get multicast routing working on OpenStack. Storage challenges, obviously Gaurav was talking about all these unique storage requirements for running video applications on OpenStack. It's pretty key. So what Cisco did to actually address the storage challenges, we'll talk about that. Complex multi-VM with all the service level agreement. Video applications come with their own unique requirements. One application has a lot of memory requirements. One application has a lot of memory processing requirements. One come with storage requirements. Putting them all together in a work stack, in a workflow, and running them in a cloud environment is pretty challenging. So we had to come up with very unique innovations inside the development team to actually run it on OpenStack. Security requirements, obviously. When you deal with content, everybody knows the strict content requirements and security requirements around handling data in the cloud, right? And obviously content affinity requirements. When you're dealing with content, you have certain geographic requirements where you cannot push content from one cloud to another cloud so easily as you push other data, right? So what we did, right? So we had to actually bring in multicast routing into the cloud, into the OpenStack environment. We came up with plugins, the ML2 plugin to get OpenStack ready for multi-stack, sorry, multicast routing. And then we actually innovated in the storage area. We took the SwiftStack, enhanced the SwiftStack, added some extensions for media-ready extensions, right? So these SwiftStack extensions made us to actually work with the storage open source, object store community to get that extensions added so you can actually do video applications on top of open source, object stores. Complex multi-VM with SLA, so we use a sealometer, we use heat templates for actually orchestrating these multi-complex virtualized machines to actually integrate them, service change them together to get video applications running on OpenStack. And security requirements, obviously there's a lot of security requirements. We enhanced our cloud service router to actually run in an OpenStack environment to provide the content owners the capability to run multi-tenancy video applications in an OpenStack cloud environment. And obviously we have intercloud that Lou was talking about to scale, to scale Cloud DVR, for example, you want to scale Cloud DVR at some point from one data center to another data center, you can interconnect the clouds between together using intercloud application that Cisco is coming together, federate them and interconnect them when you need scale and then scale down when you don't need the scale. So we are doing all these kinds of innovation to get video applications running on OpenStack. So some of the challenges I've listed. So now I got it. Thank you very much. I think that that's just one example we could have brought up here, mobile packet core, like we say, many of the network function virtualization areas. There's a lot of different parts of Cisco now that is moving on to OpenStack. And what I really wanted to do with this video team was show that they really chose to use as much of the componentry they could in OpenStack without having to sort of go off and then they basically do their own. So the last topic I just wanted to discuss here is that like many companies, we have a very large developer community and one way we reach those is through the web. And so I wanted to now talk about how we've set up within the developer community, Cisco's developer community DevNet on Cisco's website, a special community for OpenStack. And what we're trying to do there most importantly is instead of trying to put all the content there is to be a repository for links back into the community. We want to always drive people back into the community, all of the contributions we've made or on GitHub and associated with it. So we're really not trying to capture content there instead make this a reference source for people to go back into the community. And I would urge other companies to use the same kind of model. This is also a place where we plan on experimenting. And so that's why a lot of the innovations that we've shown earlier are showing up on this site today. And if I can get Rohit to come up today, perhaps he can walk us through what this is. Sure. If you want maybe I can... Let's get to the browser. Yep. So we, as Lou talked about and we had a good example of a solution. So the idea of the DevNet program is to enable our customers as well as partners to find information so that they can quickly get started with some of the technical stuff that they're dealing with. So we have concept around DevCenters which are around a major technology area. So we have like example here, a cloud DevCenter that we have opened up. And within that we have built upon a specific microsite that is dedicated for OpenStack. So within the cloud DevCenter here, we have the OpenStack microsite. So the microsite here is dedicated to provide all of the contributions that we have provided within the community. So if I go into the OpenStack projects at Cisco, these include all of the plugin and drivers pointers to the GitHub repos. So this is one of the challenge that we see from our customers that great, you're making so many contributions but how do I really get to know a consolidated view of all the integrations that you have done? So the first table for example, list all the product integrations with Cisco and the second some of the things that Lou talked about the incubated projects, the AirVos, VMTP and we have a few more that we have released for example Cloud Pulse. So we'll be adding information about all of this here so our customers can go back and look at this information and make a sound decision in terms of what they want to integrate in their OpenStack environment. Some of the other technical information that we are contributing is in terms of technical briefs. So for example, Neutron IPv6, a very important feature within Kilo release. Cisco has been actively implementing a lot of blueprints within the Kilo release and the Neutron IPv6 tech brief which is publicly available captures all of these information that we have contributed and it goes through technical details of what every feature implements within Neutron. This is something again we have worked with the community and with the Cisco OpenStack IPv6 experts. Actually a lot of this content is already getting reused within the OpenStack IPv6 Wiki page as well as the Neutron networking manual. So all of that is going through the Git review process within the community. So again, a very good example of how things that we are building for our customers as well as for the community is getting implemented and getting included within the upstream community as well. Going back to the side, again, from I already talked about the downloads, Lou talked about some of the customer events as well as Cisco Live. So we have created this video catalog where we are providing all of the sessions that are contributed by Cisco and participated by Cisco members for OpenStack related talks so that we have a reference for other folks to look at it. So in summary, what I wanna say here is that through this DevNet evangelism program, we wanna make highlight Cisco's contributions from a technical point of view as well as from a product point of view. Stack Analytics and other sources within the community provide you that information. But a lot of times it's very hard to go through all that sea of information and we wanted to make sure that developer.sysco.com slash OpenStack is your one point portal to find out all the involvement areas for Cisco with OpenStack. Great, thanks. So that's why I'm really pleased to be launching this today here at the summit and I hope this becomes a very valuable resource for all of you. So I hope what we've been able to talk about today is something about how a company such as Cisco has been able to adopt OpenStack and drive it into many of our different businesses and I think we might have time for one or two questions. Anybody's trying to do something similar in their own company and what are the challenges? Yeah, oh, you had a question earlier, yes. There's a mic over there because this is being recorded. You mentioned earlier that Cisco worked to help Comcast deploy OpenStack. I was curious in what capacity was it that DVR, cloud-based DVR? It's actually what they're showing today which is their X1 Xfinity application that's running on top of OpenStack. And so we help them set up OpenStack, get them involved in the community and a lot of the education and training and technical support around that so that now they're a full-fledged member and we're continuing to work with them on a number of different initiatives. Yeah, which team we're seeing Cisco is working on the development of the group-based policy with OpenStack and the L4 to L7 functionalites that are working on. Yeah, so actually the group-based policy is done within a small team that's associated with what we had as our assembly acquisition and that that is developing a lot of that. They actually are one floor above where my group is and so we have a very close interaction with them and then we also have within my team a lot of the services interfaces that are going into service orchestration and things that are going into Neutron itself. But I think those of you who are interested in group-based policy, I think that will be a very interesting area to watch over the next couple of years because it's a much simpler way to approach scalable applications where you really are trying to express policy and policy is I think the next thing in software that we really are trying to find ways to use policy to drive these things so we get out of configuring things and setting up ports and everything else which it just caused for Arabs. Yes. Hi, I work for a huge telecom in Brazil and we approached Cisco back in 2013 and we were presented with lots of technologies from Cisco. Basically, we bought a bunch of Vblocks, Cisco Nexus and stuff. And still today, if you take a look, if you contact a Cisco salesperson or if you take a look at what Cisco is doing in terms of marketing, if you don't pay close attention, probably we're gonna end up having an idea of Cisco being MetaCloud, probably a tail-f. Still, what do you guys, your particular group that you mentioned, are doing in order to get Cisco as a company aware of this multiple things? Yeah, I mentioned any large company, as I'm sure you're aware, we have a very large field sales force and educating particularly around the globe has its challenges. But I think for everybody here, it's important to understand where Cisco's emphasis is. First is in terms of driving the overall open-stack community. We very much believe in this platform and where it needs to go. So we're, and that's where my group is a rather substantial group within Cisco, we're driving a lot of those core contributions back into that. Out of that team, we're also working with each one of our, with a smaller set of customers, like I'd mentioned with Comcast and others, where we are having direct engineering to customer engagements, because open-stack is very difficult sometimes for customers to do on their own. Then we have the inter-cloud initiative, which is where we are setting a global network of open-stack clouds. And our meta-cloud acquisition was an interesting one because we recognize that for many companies who do not have the engineering expertise in-house, they would prefer to basically have a private cloud as a service. So have a company such as Cisco come in, build a private cloud, Cisco manages it and operates it, and the IT organization just uses it. It's like having their own captive IA, you know, AWS in-house. And so those are the three areas, really. One is in terms of the contributions in working with our customers. The second is in terms of inter-cloud, and the third is that private cloud as a service today. Okay, two more minutes, so maybe one more question. Or if not, I know the room is really hot. We can, okay, let's do this. I will be here right outside in the hallway, as will the other presenters or whatever, and I appreciate your attendance today. Thank you very much.