 All right, can everybody hear me? Ha ha ha, very funny. Welcome to the session about OpenStack and VMware. So my name is Dan Wellent. I head up our product management efforts at VMware around OpenStack. So prior to this role at VMware, I was the PTL for the OpenStack networking project, formerly called Quantum, now called Neutron. I won't tell you which one I like more. So anyway, my goal with this talk is really to address a couple things. But one main point, if you only walk away with one thing, I want you to walk away with a sense that even though a lot of people in the press or certain companies may try to kind of give you this impression that there's a fork in the road and you either have to go the VMware route or the OpenStack route and never shall you meet again on those roads. I want to help you understand that this is really not true. And of course, when I say that right now, it probably sounds like a marketing platitude. So pretty much most of my talk will be trying to convince you, I give you very concrete evidence about why VMware is legitimately interested in investing in OpenStack and helping our customers be successful with OpenStack. So I actually think a lot of the misunderstanding around VMware's relationship with OpenStack actually has a root and a misunderstanding of what's interesting about OpenStack at all. I think the easiest thing for someone to write an article about on OpenStack is to say, oh, OpenStack's free. The interesting thing about OpenStack is that it's free. Free is in beer. And my personal opinion, and I think a lot of people who spend a lot of time with OpenStack, so this probably isn't the interesting thing about OpenStack. A, because there were other free open source clouds before this. Things like Eucalyptus CloudStack. But OpenStack's way bigger than them. Similarly, no one who's actually successfully deployed OpenStack would ever tell you that it's free. You've made some significant financial investment somewhere along the lines, either building your own team, or hiring help, or buying software licenses. So what I really think is important about OpenStack is free dumb. And what do I mean by that? Really, OpenStack gives you extreme freedom of choice. OpenStack is a framework with a common set of APIs that then lets you choose what compute, network, storage, and management technologies you want to plug in. OpenStack also gives you choice in terms of how you want to consume that cloud software. It lets you do everything from building up your own team of developers, who know OpenStack inside out, pulling the code, building your own packages, and pushing them into production, all the way to buying a product from a vendor where you're paying them software licenses to deliver OpenStack to you. And there's a whole spectrum in between. So both the choice of technologies and the choice of delivery mechanisms I think are actually what's most interesting about OpenStack, because it gives you that freedom. So with that in mind, really, when you make a decision to use OpenStack, it's really only the first of many choices that you need to make. Really, what you as a cloud architect or cloud admin need to do is then say, OK, what software and hardware technologies are going to give me the features that my application developer needs, the scale that I need, the performance that I need, the management tools and troubleshooting capabilities, the ability to provide my users with SLAs if they need it, those types of things. And so depending on the cloud, depending on the type of users you're trying to serve, you're probably going to come up with different answers for that compute, network, storage. And so our goal at VMware is to convince you that there's value in providing VMware technology in your OpenStack cloud and convince you that VMware provides differentiated value from everything else out there. This is no different than what any other business does every day. And VMware is committed to helping customers with OpenStack deployments, either in terms of just consuming an individual VMware technology. For example, you could have NSX that runs with KVM hypervisor out there in Seth Storage. You could consume multiple components of VMware technology, maybe a hybrid cloud that has some KVM, some vSphere tied together with NSX. Or you could go whole hog and take the entire VMware product portfolio and run OpenStack with it. Again, that's the freedom of choice we're talking about. So a lot of people still, for some reason, don't believe that we're really embracing OpenStack at VMware. And so the argument I often tell people is that you can afford to embrace customer choice and embrace things like OpenStack when you really are confident and think that you have really solid and innovative technology. So starting with that is our assumption in terms of defining our OpenStack strategy. The rest of what we're doing, in my mind, seems incredibly obvious. Well, if you have great components, what are you going to do? You're going to contribute code to OpenStack to make sure those components are well integrated into OpenStack. Well, then you've got customers who are trying to be successful with OpenStack in your stuff. And well, you're going to help those customers out, whether the problem is with your driver or with OpenStack in general. So you're going to contribute overall to OpenStack to improve the platform to make it easier to use. And you're going to talk to your customers and help them understand why your product provides differentiated value. Again, this is what any other company that's trying to deliver OpenStack is doing as well. VMware is no different from that front. And as you'll see later in the talk, we're working very closely with members of the OpenStack ecosystem to make sure that it's very easy for you to take advantage of OpenStack on top of VMware technologies. Even if you think NSX is the best networking technology, well, if it's a total pain in the butt to use, and something else is easy, you're probably going to start out with that something else. So we're also investing in making sure that it's easy for you guys to consume OpenStack on top of VMware. So the way I sum this up internally, when I'm talking to our OpenStack engineers, it's very simple. It's to say, we want to contribute to OpenStack. We want to improve our products, make them worth very well with OpenStack, such that we have a strong argument that the best way to run OpenStack is to run it on top of VMware technologies. It's very simple. And I still can't for the life of me figure out. I think some degree I think it's just the press people love a versus argument. They love that story of VMware versus OpenStack. But this is the reality. This is the strategy VMware's acting on is. I think it's pretty straightforward. So let's go into a little more detail about what this means. So on the upper side of this picture, you see the OpenStack projects you're familiar with. You have Nova, Neutron, Cinder, and Glance. And then you have the tools that you're familiar with for accessing them, things like the Horizon web portal, CI tools, or some automation scripts built against the OpenStack API. And this shows all OpenStack underneath. But as I mentioned before, people may consume individual components of VMware technology or a full VMware stack. But Nova can plug into ESX and vCenter. Neutron has a driver for NSX. Both Cinder and Glance can interact directly with VMware data stores. And then the set of cloud operational tools that VMware provides can manage those components and help you with things like troubleshooting and capacity planning. So again, to a user consuming cloud resource is what they see is Horizon. They see the CLI tools. They see the OpenStack APIs. It is OpenStack to them. But it's the set of technology decisions on the bottom you've made is where you may or may not choose to use something like VMware. And then just one thing to make this a little more complicated because it wasn't enough. There's also a technology called VCAC that VMware has. Think of it as kind of a cloud broker. But we was already able to consume resources from vSphere from AWS, from different pools of compute network storage capacity. And we just recently announced that it's now able to consume OpenStack APIs as well. So this is another example of how basically our strategy is to make sure that if you invest in VMware technology and then at some point in the future or right now decide to go OpenStack, those investments still continue to provide value and provide differentiated value within your OpenStack deployment. So my mom always just said the proof is in the pudding, but that doesn't have any meaning here. So what I really say is when people ask me about how serious VMware is about coding, I say about OpenStack, I say the proof is in the coding. So if you guys have seen Stacklytics before, it's a site run by Mirantis, really cool way to look at what's actually going on in terms of OpenStack contributions. A year ago, VMware applied to be a gold member of the OpenStack Foundation. And a lot of people said, oh, don't let them in. They're not really going to contribute. They're just saying it. They're in here to undermine OpenStack. And what you've seen is us progressively not only continue to maintain our investments, but actually increase those investments. Not just in absolute terms, but actually in relative terms to everyone else as well. So with the number seven overall OpenStack contributor to core OpenStack projects, now that individual number isn't important, or who we're ahead of or whatever. The point is that if you look at the OpenStack ecosystem, we're of 100 plus companies. We're way at the top in terms of the overall contributors to OpenStack. What does that mean in practice? In Havana, we had 17 developers contribute code. Some of those contributed a lot. Some of those contributed a little, 17. We had 319 different commits. And the one I'm most proud of probably is 3,693 code reviews. And the reason I'm so proud of the code review number, that's actually number five overall, is because code reviews represent you actually helping the community out. Because most code you review isn't your own code. So this is another example of how VMware is not just in it for themselves. We're contributing to the community. We're improving the overall quality of the OpenStack code base. So another question is people get to say, oh yeah, you're contributing a lot, but all the contributions are a nova. It's all those NYSERA guys. I was a NYSERA guy, so I'll take a little credit for that. But the reality is that we're nearing a tipping point. So I tried to do kind of an infographic where the size of the font maps to the numbers of lines committed. So you can see the majority, just barely, of our commits are still in neutron. But we're very quickly revving up our Nova and Cinder commits. And we also have commits to other projects like Docs and Tempest and DevStack. So I definitely think that by Icehouse we'll actually have more commits outside of the networking side in compute and storage from VMware than we have on the network side. We'll continue to do our best to be the number one contributor to the neutron project. We're not abandoning that, just saying we're putting so much additional energy behind our OpenStack efforts that I think overall Nova and Cinder will catch up to where we are with neutron. And of course, we can't do that all on our own. A very important part of OpenStack is the community and the developer community in particular. So I wrote some scripts and figured out there's just about 200 people who've reviewed a VMware patch in the Havana release. So my first idea was let's get them beer. But then I thought of logistics of trying to get beer in Hong Kong and I said that's not a good idea. So I got the next best thing, which is a Starbucks gift card. There's a Starbucks right upstairs. And I figured this will fuel a lot of future late night reviewing sessions. So it's a good investment for the OpenStack community. So if you or anyone you know has reviewed a VMware patch, just go to our booth, show your badge. We've got a list of the people. And you'll get a free Starbucks gift card. So where has all these commits and reviews gotten us? So I'm going to talk about kind of where we are in terms of Neutron, Nova, and Cinder. So in the Grizzly release, which was six months ago, and Havana release, which is the one that just came out. So Neutron, as I said, is the project we've been deeply involved in for a long time. So we considered our Grizzly release of Neutron and the MVP and NSX support in their production quality. And essentially what happens is as each new OpenStack release comes out, we just fold additional features in there that the community is working on and that we have support for in our product. So things like firewalls of service, low balancers of service, those types of things came out in the Havana release. And we have implementations, both community implementations and implementations that use a VMware NSX networking. On the Nova side, as I said, kind of our public commitment a year ago was to add vSphere support to Grizzly. We did that. That meant that you can use vSphere with advanced features like vMotion, HA, DRS within OpenStack. So I kind of considered that code that landed first in Grizzly to be kind of an early beta and we'll be talking about a couple of customers who were brave enough to dive in and try it out and we'll talk a bit about their experiences and their feedback soon. And then effectively what we did for Havana was we really just kind of used that feedback from those customers to fill in gaps, to fix bugs and really document the heck out of this whole setup so that more and more customers can be successful with it easier. And so we'll talk about some initiatives around that toward the end of the presentation. And then Cinder, this was the big new thing that landed in Havana from a VMware perspective. So in Grizzly, there was really only support for a very basic iSCSI driver. And what we added was a vSphere data store driver which we'll talk more about later. But it basically means any shared storage you can use with VMware, with vSphere, you can now use for OpenStack Cinder with essentially no additional modifications. And I definitely want to call out that particularly in the Nova space, we got help from folks like HP from Canonical and IBM. They also contributed to the vSphere driver at Nova. So with that, I want to take you through a couple scenarios, highlight a couple customers who we've been working closely with. This isn't to say these are the only people deploying OpenStack on vSphere, but these are people who we had existing good relationships with and have kind of been helping out and getting feedback from, trying to help make them successful with OpenStack on vSphere. So I have this picture up here because in my mind they're like the pioneers, right? Someone had to go out there, someone had to rough it, someone had to deal with the bugs and the feature gaps to smooth the path out for the rest of the people coming later. So the first one, if any of you had a chance to go to the talk yesterday is PayPal. So they've been quite public that they're already running with OpenStack on KVM and they're looking to get vSphere assets under that OpenStack umbrella as well. So I won't go into too much detail because if you really wanted to know a whole lot about this, you can go to the talk or find Scott Carlson who's here as well. Is Scott here at all? I'm trying to think. I don't see him, so. But anyway, he's a great guy. They've been a great advocate for us and they've been very helpful. They've identified some key gaps like config drive support. You know, they've helped us iron out some reference architecture stuff. So PayPal has been a great advocate for us as well. Another person we've been working very closely with is Intel and since DOS is actually here, I figured instead of me trying to fumble through telling about it, I would just hand it off to him. Well, thanks Dan. So I'm DOS Cam Hout, I work in Intel IT. I'm a Principal Engineer and I cover basically all of our cloud activities from software as a service down to infrastructure as a service and just to quickly give you some history. My background is running our design grid. I was basically born at Intel in 97. We have about 60,000 servers across 60 plus data centers and we call this clouds basically uncle or mother. So it had a lot of the tributes that we'd expect from a cloud self-service for the end users, API that everybody can consume and they can use it on demand. Back in 2010, we felt we were early adopters of the concept of enterprise private cloud. Right now we're running about 13,000 plus VMs across 10 data centers and 75% of all of our enterprise service requests come through there. So people don't actually have to talk to IT, they go to their self-service portal and this is all based off of basically VMware technology and then we had our own proprietary code that we put on top for workflow orchestration in a portal because when we did this in 2010, there wasn't anything. And then in 2012, we stood up our first open source private cloud. So if you can imagine our design grid is heavily focused on open source. We thought it would be good to move forward with some open source components. But as Dan said, this is all about the control plane, the APIs that make it a uniform, whether it's private or public that our developers can use against. And right now we're about 1.5,000 VMs across two data centers. And the key point here is this is running cloud aware applications. So you hear people talk about the pets and the cattle. So all enterprises have a lot of pets and they're gonna be there for a really long time and we need to take care of them and they need resilient infrastructure. What we focused our open source private cloud on was initially was specifically on the cattle concept of things that can die that are built for a failure. So another key point is on the convergence. So we have large silicon design. We have validation labs and we have enterprise hosting. All these are very, very different use cases but we've chosen to use OpenStack as that API control plane for all these different use cases so that our developers that are doing all these different types of things whether it's designing chips or just building web apps that our end users inside of Intel can utilize, they have one way to communicate with the infrastructure. And what we really love about OpenStack and the reason that we've seen all the enterprise vendors step forward and build plugins is that we can run our existing infrastructure. If you can imagine an enterprise IT shop like ours is pretty large. We have a pretty massive scale and we're not just gonna throw everything away. That makes no sense. And as well, there's features that we love in a lot of these products. So we can run our existing infrastructure and our new infrastructure all with one control plane for our end users. So just real quickly on current stats and plans. So we've been working with Dan's team so he's been gracious enough to spend a lot of time with us and help us out. We've been running a POC. Looks pretty actually similar to PayPal's. They're a little more further ahead than us. Controlling vSphere through VC. And what we are doing, this is an enterprise app platform. So it gives us live migration. It gives us the capabilities that we need. We need to be able to see live migration DRS, VMware HA and the future tech investments that VMware is making that makes their platform very strong in the industry and something people have been using for a long time. So what's working? We're doing VM orchestration. We're doing this on Grizzly today. So we're pretty excited about stepping forward into Havana. So we love all the many fixes that are happening. We have thin provisioning, multi-clusters. The ability to control the data stores is a huge thing because most of us don't use actually iSkezzy. So having the ability to use VC to talk through the data stores is massive. OpenStack of controlling all the existing virtual center deployments. So we actually want to control everything. We wanna get rid of our proprietary code that we built internally, our technical debt and allow OpenStack to control this. And big call out to Dan's team to continue their support. We saw the contributions go up and our team will probably even start coming and helping out to make that real. And last point is federated experience. We run in a public-private model today. We are hybrid. I think every enterprise will be hybrid. And we wanna be able to see this as a control plane that we can use whether it's a public environment, whether it's a private environment across a wide set of solutions and multi-platforms as the consumption model. So that's what I had. Thanks, Dan. Yeah, it would have been pretty hard for me to say all that. Cool. Isina or Justin here? I think, oh, there's... Oh, that's right, that's right. So I also, you know, another person, in fact, remember, you know, at the summit six months ago I was helping them debug an issue on OpenStack and vSphere, I think. His app tier, these guys are great. They're in Australia. They both have run an internal OpenStack cloud and help people deploy OpenStack. And, you know, VMware is, you know, by far, by far the dominant platform for a hypervisor in Australia. So that's why they're so excited about OpenStack. So they've been great in terms of providing feedback. I don't want to steal too much of Isina's thunder, though. He's giving a talk later today talking about private cloud in Australia for OpenStack and VMware. So that is at 340 today. So definitely go there. He's given me some feedback, but I don't want to steal his thunder. So definitely check that out if you're interested. And the last one I want to talk about is Serpro, which is a company I never heard of until a couple months ago. They're actually the IT arm of the Brazilian government. So these guys are massive, right? And, you know, they basically had an internal initiative around OpenSource and, you know, someone in the org had decided we're going to go OpenStack. You know, OpenStandard is the right thing. Let's go OpenStack. And so they started implementing OpenStack. And they called us about NSX. And we came in and kind of showed them how it worked. And they're really excited. And then we found out that, like, everything else in their infrastructure is VMware. And I said, well, why are you running KVM instead of VMware here? And they're like, oh, we didn't even know VMware was possible in OpenStack. Are you kidding me? If we can do that, we're totally going to do that, right? And this, in a lot of ways, highlighted the fact, you know, I'm really excited about the opportunity of OpenStack on vSphere to accelerate OpenStack into the enterprise, right? There's a lot of people who really know how to operationalize and are very comfortable with vSphere already. And so, you know, this was an example. We went from POC to production, I mean, a small production, but their production with like 80 different tenants serving up, you know, a basic app, right? That they replicate for those 80 different tenants, right? And we did it in like under a month, right? Because it was that simple, because they already knew how to run production vSphere infrastructure. All they needed was the OpenStack control plane deployed on top of it, right? And then they could just provide people with access to the Horizon GUI. So yeah, that was a, I like that example because it was just, oh, let's go, let's get started. And we were done in like literally about a month. So next up, I'm going to talk a bit about, you know, why we think VMware technologies within OpenStack are compelling. Now, this is obviously part sales pitch, right? This is my job to educate you about the cool stuff we're doing, right, within OpenStack. But the high order bit you should take away from this is just that there are many, even though it's OpenStack, right, it's, you know, it's important to think about what components you plug into OpenStack and what are the trade-offs between different components. And so we'll highlight some trade-offs, some things that our platform has, and that may or may not be relevant to your OpenStack cloud. But this is kind of what I meant about that second order decision you need to make after you decide to go OpenStack. So first, let's talk about Nova and vSphere. And so I don't really need to belabor the point here. I think a lot of people know vSphere's reputation as a leading hypervisor out there. DOS talked about, you know, the availability features like vMotion and HA for, you know, for those kind of pet workloads that can't handle their own failures, which is obviously the majority of workloads today. You know, but there's also a lot of, you know, really cool stuff within vSphere about resource management and resource protection, protecting against noisy neighbors, letting DRS and SDRS move workloads to avoid hot spots and better use your existing capacity. And so if you actually look at this, I think one of the things you often hear is that, oh, well, yeah, we need vSphere, but we only need it for our, you know, our old apps. I think one of the interesting things here, if you look at, you know, the concept of reliability, security, performance, right, those things are all relevant, regardless of whether your apps handle failures or not. Same things like DRS and resource protection. So, you know, I think they're definitely compelling technical reasons to continue to use vSphere in certain deployments, right? And then there's, you know, beyond the technical, there's probably even bigger of the operational benefits, right? So this is an enterprise grade technology, right? People have been deploying this in the enterprise for a long time. You know, it doesn't have the sharp edges that other platforms may have, right? A lot of enterprises already have the expertise for deploying vSphere. And like I gave the example with SerPRO, in a lot of cases this can really simplify OpenStack adoption, because it's just about deploying the OpenStack control plan services at that point on top of vSphere. So the next one to talk about is a networking side. You know, I think you've probably, probably heard the word network virtualization being bantered around here. Thank God people are finally talking about that, not just SDN. The buzzwords keep changing, but the basic idea of network virtualization, and this is something that we at NYSERA pioneered, is that you want to be able to create virtual networks on top of the capacity provided by a physical network just the way compute virtualization creates virtual machines on top of the capacity provided by a physical server. So what this means is people can programmatically ask for private networks and load balancers and firewalls and no one needs to go out and change your physical network. It can happen on demand because it happens by processing within the vSwitch, which we reconfigure programmatically. And you know, like all things, right, the details here matter, right? It's one thing to get it up and running in a lab and it's another thing to re-running it in production at scale. So things like tunneling performance, how high availability is handled, how things scale out, as you go beyond the capacity of a single server, all those things are important things you'll have to consider when choosing a driver for neutron. One of the cool things, there are two of the cool things about NSX is that it's multi-hypervisor, so it supports KVM, Zen server, and ESX. So as I mentioned, we have lots of deployments of NSX that don't have vSphere at all in the picture, right? They just see value of NSX independently of the hypervisor platform and want to deploy it. And then also we have a really broad partner ecosystem. This is something VMware is typically pretty good at, right, in terms of working with hardware switch vendors and hardware load balance in companies to be able to integrate their physical switches and also allow those switches to be controlled by NSX as well. So this is something that you'll see coming out in the next couple of months and I think is a really cool big step for the NSX platform. And of course, the operations is always kind of the ignored thing in OpenStack. So big benefit of NSX is the tools it gives you to really have visibility about how these virtual networks are being created, right? Is there a failure in the physical network that's affecting this virtual network? Those types of questions. The last one I want to talk about is the vSphere data storage driver. So in DOS I already kind of touched on the core value of this driver, which is that it lets you take the storage you've already carved out for your vSphere, right? And just consume it at center capacity, right? Without having to expose iSCSI directly to your Nova compute nodes or anything like that. So it really simplifies the process, but at the same time it lets you leverage all of the work that's been done on the vSphere side. You know, VMware and storage vendors have been working very closely for years to make sure that their solutions are really well validated. They've worked on making sure that things are like snapshot and clone are really well accelerated. And when you use this driver you actually get all of that for free. So that's really cool. And then another really neat part of that is that the same driver enables the use of a new technology from VMware called VirtualSan. So VirtualSan basically gathers all of the hard disks and SSDs that are in your hypervisors, right? And oh, I believe we have a cool animation. Oh, Valence, and did you do that? Right? And it pulls them together right into create this kind of virtual array. So you can treat it exactly like you have a dedicated physical storage array, except you put this together with low cost commodity devices. We do the replication appropriately so individual devices can fail and all of that, right? And then the other really cool thing is the amount of SSD that's used for an individual VM is configurable. So for example, if you have a database server that really needs a lot of IOPS, you can give that actually more SSD cache so it can get much better performance and you can do this at a per VM granularity. So it's actually a really exciting technology. We're gonna continue to do work in Cinder to make sure more and more of the cool things about vSAN are exposed in Cinder. All right, so let's check how we're doing on time. I wanna see if I... So we have a live demo that we will give in a demo theater at, I think it's two o'clock. Is that right? So let me just quickly pull up. Let me see if you, can you guys see this all right? I think that's all right. So I'll just do a really quick example of what we're looking at. So this is obviously the main OpenStack interface. Again, when you're a tenant using OpenStack on VMware, this is what your tenant sees. Same thing, right? It's the standard GUI horizon of horizon or the CLI tools or you can write API scripts. It's all the same. But then if you're an operator, for example, you still get your interfaces that you're familiar with to help troubleshoot problems or do capacity planning or understand what's going on in your environment. So for example, here we have, this is a pretty simple vSphere cluster and you can see latency from Hong Kong is not great. But you can see, we've got, we've got, what is it about? Gotta scroll over. So we've got about 600 gigs of RAM in this cluster. And so the VMware driver basically gets pointed at clusters as a whole. So that individual VMs can be moved and migrated by things like VMware HA or DRS without it confusing Nova. The other interesting thing we have going on here is we have a vSAN data store. So we have a bunch of them. We have some traditional data stores but we also have this vSAN data store. So again, we're gonna be provisioning VMs that will effectively have shared storage. We'd be able to live migrate them. We have DRS enabled, right? But there's actually no shared storage right there. It's just the storage in the end host being clustered to look like shared storage. And then the last thing we have in the mix is the NSX manager. So this is a GUI, again, talking to a scale out control plane for your network. And so when I go boot a VM here, it's gonna end up talking to both vSphere APIs to create the VM. Oh, that's great. Let's see. And then it's also going to talk to NSX to create and provision the network. So let's just go, I love the tiny windows you get when you change resolution. Thank you for the reminder. So I'm gonna, with NSX, I can pick as many networks as I want. I could go create a new network if I wanted. But in the interest of time, I'll just use an existing one we created, right? This is launch. So you can see, this looks exactly just like, you know, the same feedback a user gets, right? That's the point of OpenStack. It's a uniform API regardless of the backend technology. So again, the user would only ever deal with these APIs, but we can swap over and say, well, what if an administrator wanted to troubleshoot this VM? So one of the cool things we've been working on, actually, is, so what was the name of this one? It was Hong Kong One. So let me see. Once it's provisioned, I can actually come over here. Let's see. I can search for it. I can search over the OpenStack identifiers, right? And those identifiers are pulled into vCenter and they're meaningful there. So I can then click on that VM, right? And I can see everything about it. You know, so I, as an administrator, get my comfortable interface for troubleshooting. I plug into any third-party tools that I'm already used to using with VMware. Like, we have VC operations hooked up in this setup. And so I get my kind of comfortable VMware interface while your tenant gets the standard OpenStack interface. So you can see this VM is now up and running. You can see that it's got a hard disk and we should see here that it's actually on the vSAN array. And then if I wanted to do something like go over and create a Cinder volume. So if you notice right here, and again, this is one of the administrators, you see, there's only one hard disk connected to the VM right now. But we're gonna use the VMDK driver to go create a volume and attach it. And let's make that 10 gigs. We will attach it. How's it gonna make me specify? So what it's doing in the background, it's actually allocating the space on demand. So in our, in the VM or VMDK driver, it doesn't actually do an attach and tell you attach it for the first time. Because then it's optimizing where that VMDK is placed. And if a VM moves, or for example, you attach and detach it to another VM and there's a better location to place that VMDK, it'll actually get moved to that location over time. So that's what SDRS does. So again, it's optimizing the location of your data storage based on the placement of the VM. So if I go over here, once that attaches in good shape, I should be able to refresh here and see now we see a second hard disk here, 10 gigs, just the size of the volume we've created and also on the vSend data store. So here you've got local storage where you can still migrate the VM, right? You've got a volume that you can attach and detach to any VM in your deployment. And it's all done without any shared storage. So I think given the time, I'll probably skip on going into the NSX stuff, but you can stop by the booth and get a deep NSX demo if you want. We can basically show you some of the troubleshooting tools is what I was gonna do, pull up NSX, show how you can see what tunnels were created as a result of that VM, see the result of our probing of the physical network to confirm that those tunnels were up. So again, you can imagine the difference from an operational experience if you're able to access kind of rich management interfaces like this when troubleshooting issues, when capacity planning, et cetera. So I'm gonna swap back over to the slides. We've got a couple more things to wrap up. So one of our goals is to, like I said, make it really easy for people to get experience with OpenStack running on top of VMware. So this is something we launched a couple months ago. My goal is just how can we eliminate all the possible hurdles and make it as easy as possible for people to try out OpenStack on vSphere? So we created something called Vova, which is basically a single OVF that you can download. And so if you have an existing OpenStack, or if you have an existing vSphere cluster, what you can do is basically open up the file dialog, say deploy OVF, put in a URL, press go, a dialog will pop up. It will ask you a couple questions. What's your VMware cluster IP? What's your VMware, your vCenter username and password? What's your data store? What's your cluster name go? All right, and then it will deploy and you'll have a real OpenStack on vSphere deployment running in your environment. So this is basically taking out, taking all the control plane services of OpenStack and running them in a single environment. All right, so this again, this is not for production. This is just a toy for you to learn, particularly because there's a lot of people who are very familiar with VMware and want to learn about OpenStack. So that's who this tool was geared for. So you can check out the URL there, download it. We just did an updated release today. So it includes that kind of vCenter OpenStack awareness that I just demoed. It also includes all the bug fixes from Havana. And then in October, we're gonna come out with another version that also lets you use it with NSX if you have NSX access. So right now it's Nova Network only. But we just kind of keep wrapping this. And again, the model here is not that this is something you'd ever use in production. It's just to help people learn. So, but then some of us decided that even that wasn't easy enough. So then what we did was we created, used some infrastructure VMware has called the hands-on lab. And so if you go to that URL, the difference is the lab at the end, right? What you can do is you can press a button and in 30 seconds, you get a remote desktop of an environment that's OpenStack deployed on top of a small vSphere environment. So you can play around. We've got a dedicated set of instructions for you to follow through. You can learn about OpenStack on vSphere, kind of at your own pace. And then you can play around with the environment and go do whatever you want to do as well if you have additional time. So we'll also, we have a session Friday at, is it 11? 11, I think, that we'll actually have a big room and have people going through this. So we'll have proctors there to help and ask questions. And if you have more advanced questions, we'll kind of, we'll get a mic and answer those as well. So, and again, kind of like the Vova, this uses Vova under the covers, right? So we'll have updates for NSX and BSAN in late 2013. So our goal is to keep kind of rolling out new modules to this lab. So it's, again, really easy for you to learn about OpenStack in VMware. So then finally, the question is, well, how do you get access to this for real production deployments? So, you know, we have a set of partners. We've already announced a partnership with Canonical back in April, coordinating around vSphere and NSX. They're giving a talk tomorrow at 9 a.m. That's gonna talk about how they can use Juju with Ubuntu to deploy vSphere and NSX. We've been partnering with Red Hat around NSX specifically, working with joint customers to make sure the OVS modules that need to deploy with Red Hat are supported by Red Hat. We're also coordinating with SUSE, right, around vSphere and NSX. Their SUSE Cloud 2.0 supports vSphere, and NSX work is something we're actively engaging with them right now. And then just today, we announced yet another partner. We're working with Mirantis around vSphere and NSX. So if you're familiar with Mirantis Fuel, it's an open source tool that you can use to simplify the deployment of OpenStack. And so, Mirantis is working with us to add support for vSphere and NSX in Fuel. So again, right, this is an example, all of these partnerships are an example of how we really prioritize customer choice. We want you guys to be able to choose a distro that's right for you, and still use VMware with it. So with that, I hope I've made a little progress in convincing you of my original goal, right, which is that there's plenty of ways that you can use OpenStack and VMware together, that VMware is serious about supporting customer choice, both in terms of choosing to use OpenStack and choosing which distro of OpenStack you want to use, right, and in helping our customers do whatever it takes for them to succeed with OpenStack and VMware. And you know, I hope I've also at least kind of piqued your interest around technologies like vSphere, vSan, and NSX, and how those fit into OpenStack, such that, you know, when you're looking at an OpenStack deployment, you consider that as one of your options. Cool, so thanks. Like I said, we've got a booth out there that you can stop by for additional demos. There's, we have a whole online community, so any questions you have, you can post them there. That's where the Vova stuff is as well. You can follow me on Twitter if you like updates on this topic. And as always, we're hiring. So come join Team OpenStack at VMware. You get a cool t-shirt. So with that, I'm happy to take questions, and these are just the other VMware sessions. So are there any questions? I think I'm just over time, so I don't know how soon we have to vacate the room. Yeah, yeah, so all these, this presentation will be uploaded. So, yeah. Haven't done it yet, because I just finished it, but. Any other questions? All right, thanks, folks.