 Looks good. Check, check. This thing's on. Beautiful. Good afternoon, I guess. It's almost noon. My name is Jason Harley. I run engineering for breakwater, and I'm here to talk to you about the breakwater private cloud appliance, what we're doing with OpenStack, and what we are bringing to the market, that we feel is a little bit different. So you may well be asking, who are you? Which is a pretty fair question at this point. We are a Canadian company. We're based out of Toronto. And we have been sort of silently participating and watching in the OpenStack community for the last two years. So first and foremost, breakwater is a hyper-converged turnkey private cloud appliance, which wins me all the buzzword bingo points I generally can get. I do well at the pub with the boys, but we have a 2U solution that has four nodes, fully HA control plane, has compute and storage, that instead of just dropping a controller on your floor, we're literally talking about installing a full HA OpenStack cloud with the initial resources to get a private cloud up and running. With that said, we're focused on delivering private cloud to enterprises. We're focused on delivering to sort of small to medium-sized dead shops, QA, or sophisticated enterprises that are ready to take that jump to fully automated infrastructure or what the industry is now referring to as IT as a service. That idea of actually delivering to your customers and not being the guy who runs IT, who has to answer my least favorite question, how come I can do it easier at home? If you run IT departments and you've been asked that question, you know how terrible it is to say because I said so. As I mentioned, we're recently at a stealth, which is sort of why this slide's entitled, Who Are You?, and we're new members of the OpenStack Foundation and Community. We're very much looking forward to participating, contributing and sharing our experiences when talking with customers about their challenges using OpenStack in a private cloud enterprise context. So why breakwater? Why are we here? Why do we think that we've got something worth talking about? We've got these five value icons. People love value icons today. And these are sort of our cornerstones or our pillars of our focus and what we're promising the market and our customers. So first of all, we like to talk about simplicity, not just simplicity in deployment, but simplicity in consumption. At Breakwater, we believe that private cloud projects are measured on consumption, not necessarily availability. If you build an amazing private cloud for your company and no one uses it, it doesn't matter how stable it is. So with that, we've built a very consumable interface. Horizon is an excellent tool. I think it's a very technical tool. I talk a lot about cockpit computing. I've been the guy with three monitors and all the screens in my evil genius layer, which was a cubicle in a building. But I think that we all use technical enough tools these days and really when we're talking about infrastructure as a service, we want to get in, get out, and get on with whatever it is that's driving the requirement for that infrastructure with our job. Breakwater is integrated. Like I said, it is a hyper-converged appliance. So we have tight integration between our software and the hardware that we run on. We also talk about integration in the context of the enterprise or the business. So we have support for talking to LDAP and Active Directory with Keystone. We also, in addition to being a hyper-conversion appliance, we will talk to Enterprise Storage. We have a scale-out object store, which I'll talk about in that next point there in the middle. But the kind of integration we're talking about here is for capacity and performance. So if you have a virtual machine that is running, say, a highly transactional LLTP workload, a scale-out object store, not the best for that. So we have some storage partners that we support and will help you get high-performance storage into your VMs. So we're scale-out. Hyper-convergence is generally about scale-out. But architecture specifically makes linear scalability of your computing storage easy. We have a fully automated way that you can just drop in a new 2U unit. It's got your four nodes in it. It's got your storage. It comes online without disrupting your current workloads. And you've just linearly expanded your storage, compute, and memory available to the consumers of your cloud, your customers or your coworkers. Time to value. This is where the analysts give OpenStack a hard time. It's referred to as incredibly complex, the permutations and combinations of the amount of different ways you can deploy it. I think approach is about 10,000. It's been a long time since I've done discrete math. We've taken a very opinionated or curated approach to OpenStack. We've done the hard work. We've watched it break. We've worked through all the problems, and we've basically decided this is how an OpenStack cloud should be built for your business. My last deployment of this cloud took 55 minutes. We went in, we racked, and the customer was launching VMs in 55 minutes. That is real time to value. No 16-week project timelines and Gantt charts. We were in and out, and they were in a position to start consuming a private cloud. Resiliency is the other big thing that the analysts give OpenStack a hard time for, rightly or wrongly. We have, as I mentioned, in our initial four nodes, a fully HA, shared nothing, controller architecture. The box will come up on its own. It will automatically cluster, bring resources online. This is the sort of thing that businesses expect. In addition to control plane resiliency, we've solved instance resiliency. Whether you boot from volume or boot from an ephemeral instance, if you lose a node, we'll bring those VMs back for you. We are welcoming the businesses pets and cattle. That is the legacy applications that take all that care and feeding, and they name Scooby, as well as the applications that most businesses are telling us that they want to build for the future. These scale out so-called Web 2.0 or Web 3.0. I'm not sure what that last one actually means, application. In addition to all of that great stuff, we're in good company. We just launched, publicly, our partnership with two other organizations who we believe have been born into the cloud. These are companies who have gone away and they've looked at the problem with infrastructure, and they've architected their solutions to do it. We are offering what we believe is the first of its kind. It's a turnkey integrated system that delivers the best-of-breed solutions of solid fire and arista networking along with the breakwater appliance. Like I said, if you need to scale out, if you want intense IOPS, I don't think there's anybody better than Solid Fire for what they can do. The switching and resiliency of Arista's data center network control plane I think is second to none. So in an effort to make Mark Shuttleworth proud, we are going to do a live demo. I've tethered to my phone, and I'm hoping that the demo gods will be good to us. So this is breakwater. Log in. We're talking to Keystone. This is OpenStack. And we now landed on the dashboard. As you can see, this is quite simple. I'm not being asked for the 50 things I might want to do. I'm not getting bombarded with dials. This is very clean. I'm in a project that has nothing. This is my sandbox. My name is Jason. It's called Jason. You can see that I have a quota assigned to me. Cool. So I want to VM. I'm going to launch an instance. I need a Windows machine for this project I'm working on, and Windows 2012 R2. I know that the theme of demos at this conference seems to be to launch a Cirrus instance, but Windows is a little cuter. And I'm going to call this NetTest. I'm going to work on the new .NET framework. One of the other things we hear a lot about business is fighting VM sprawl. I have managed virtualization environments where we were at 95% usability and we were positive that 90% of those machines weren't good for anything. So in an effort to keep business honest, we have this idea of resource leasing, if you will. You can turn it off. But let's say I need this VM for about 60 days for the project I'm working on. I can extend it easily. So I click Go. In the background now, we've done a zero cost clone from the Glantz image library. Jam that into a cinder volume, attached it to a nova instance, and knocked it up, and it's online. So that's great. The beauty of this is that your guy in Junior QA who just wants to test something and doesn't need to know anything about what a nova cinder or Glantz is and are able to take advantage of what OpenStack can offer. So my instance is coming online. I can see some details about it. I've got a small plan. It's got a network address. If I needed to attach another volume, for example, I could attach a data volume that's maybe 100 gigs, which should probably be sufficient for the work I'm going to do right now. We'll see if it's come online yet. Windows, for those of you who aren't familiar with it in the OpenStack context, it has to reboot twice. First time it comes up, talks to the metadata service, and then gets the information it needs to actually run the sysprep job, which causes another reboot. So Linux instances come online in about 25 seconds, 30 seconds. Windows 2012 R2 takes two and a half, three minutes, which I'm told is pretty good. Unfortunately, that's going to take a few minutes. So maybe to keep me honest, we're going to show an interface that people are a little bit more familiar with maybe. So this is Horizon. As I said, it's a technical dashboard, but we are OpenStack. We haven't changed OpenStack. We're committed to the Power BI program and the work that they're doing with Defcore and Refcore. So if I go into the JSON project, which is where we were just hanging out, you'll see that my net test instance has just spun up here. I can go look at it and see what's going on, and we'll see exactly the sort of things that you would expect Horizon to tell you about your instance and get a log. I think we can even get that console again. Well, look at that. So what we got going on there is back over here. Well, that instance is booting because goodness has taken a bit. One of the other things that we have done with Breakwater that a lot of sort of more sophisticated IT organizations like is resource-based billing. So what the industry calls showback or chargeback. So this project I just switched into, which is creatively called accounting, is on a monthly budget. So it's got $2,000 allocated to it. My resource pricing has been set by my cloud administrator. You can see I have four machines here and they're all doing their thing. Truth in advertising, I spun these up yesterday evening, so I've only consumed about $2.39. But if we sat here on the dashboard, these things would continue to count up. Reports are available for administrators that they can then send off to appropriate lines of business. The other thing that's a bit different when you are in a project that's in budget mode is that all of a sudden we're showing dollars back to your IT department. This can start to cause some very interesting conversations. All of a sudden internally, when someone is asking you for something that before you've just had to roll your eyes and say is unreasonable, you can actually show them the real dollars and cents that that costs your business. Again, we're very focused on private cloud meaning that a lot of more sophisticated organizations are looking to do for some projects. Worth mentioning, if you have a project that's in quota mode, let's say you don't want to show your Tiger team of developers how much money they're actually costing the business, your administrators can still generate reports that will bring back dollars and cents into your accounting system. Finally, our Windows instance has come online. And as you can see, this is about as real as it gets. I'm an admin. If I could type, I would be dangerous. Okay, so our Windows instance on our demo environment are secured by default. To practice, we recommend to customers so we don't have default passwords. The administrator account you log in and you assign a password the first time. So this is Windows 2012 R2. For those of you who haven't seen it, it virtualizes very well on OpenStack. One last thing before we break for some questions. I think I'm pretty much at time. We have an administration section that sort of gets into talking about the hardware. We also have the obligatory pretty graphs that people like to show in the boardroom. My CEO is laughing at me over there. This is what we feel is the easiest way to let administrators get into using an OpenStack private cloud. We have talked to people who are interested in breakwater as a stepping stone to a major OpenStack deployment that they want to build themselves. This is exactly the kind of private cloud that they want to consume. So with that, does anybody have any questions, concerns, comments, tomatoes? Yes. What's our core innovation in the company to repeat the question from the gentleman? I think as a team we're very proud of the way that we have done that shared nothing control plane. It is painless. There's no make sure you turn on the one in the bottom left hand corner first. We are also very proud of the way that we are trying to make OpenStack consumable. One of the core beliefs is that private cloud projects are measured on consumption. So if you build it and no one uses it, you've failed. That's a big word. No one likes to use that very much anymore. But I think that answers your question. Any other questions from the audience? If you're shy and want to talk to me afterwards, that's fine. Thank you very much for your time. I appreciate it. Again, I'm Jason. I run engineering for breakwater.