 Good morning everybody. Nice to have you here for our session. We're going to have 20 minutes just to talk about delivering OpenStack Cloud that's fit for an enterprise environment. My name's Mark Smith. I'm from SUSE. This is Simon Briggs also from SUSE. He's going to do the brave and clever piece later, which is the live demo of some of what we're going to talk about. What I'd like to do to start with is just share with you some market research of IT professionals' opinions about OpenStack and Private Cloud. Just to validate some of the things that we heard in the keynote sessions that were on yesterday. And a couple of things that we just wanted to call out. First of all, the popularity of OpenStack as a private cloud platform. Well, I think we all understand that. In our research, 81% of large corporations were either moving to OpenStack or had already moved to OpenStack. That's a big number and it gets echoed from other community studies and even from the user survey about how popular OpenStack is. Why is it so popular? Well, it's about saving costs. It's about agility. It's about innovation. What we also found is that 96% of IT professionals were now at the point where they were prepared to trust their mission critical or business critical workloads to a cloud environment and their preferred choice was a private cloud. Now, that's not always the context that we talk about OpenStack in. We talk about it more as being a development platform for new web-based workloads. But as it matures, it's becoming more than that. So just let's take a moment to talk about IT challenges in almost all businesses. What are we having to cope with? What are we trying to achieve? You will recall in the keynote yesterday, Gartner spoke about bimodal IT. They spoke about OpenStack being perfect for a faster response, about innovative new workload web-based development. Well, that's all about running faster, fitting the needs of the business, dealing with customer requirements, market and competitive challenges, being quick to respond to what's happening around us. Data center transformation is actually about what happens to the bulk of our IT. How can we gain greater efficiency with what we have already invested with? Now, not necessarily the heartland of what you would expect from an OpenStack deployment, but actually it fits really well into that. We're going to talk a little bit about fitting OpenStack into an enterprise environment. And the thing that all of us are trying to do is control budgets. We have to do more with the same or more with less. How many of us have a blank check for our IT projects? Probably no hands, right? Certainly we don't. So we're trying to do all of those things. OpenStack helps us with all of those areas, but I want to just spend a moment telling you a little bit about SUSE or OpenStack Cloud and why it's different. So first of all, with OpenStack, you can do it yourself. You can take the upstream code. That's a good thing to do. We're certainly not wishing to discourage anyone from doing that. That's about learning and developing skills. It can be complicated. Most of us in an IT world as users, for instance, if we buy a car, we don't normally build it ourselves, do we? We want to drive it out of the showroom. We want to use it for what we need it for immediately. And that's driving from one place to another, enjoying the ride, getting some benefit from it. So normally we go for a pre-built solution. We look for quality, reliability, performance, top-class service, maintenance, support, and excellent value. We want to get the best from our investment. Well, that's really what SUSE or OpenStack Cloud does for our customers. We pre-build, harden the code, get you up and running really quickly. A typical kind of conversation I have with customers is about they've tried it themselves. They found it complicated. They come to SUSE, we get them up and running quickly. They start developing and deploying workloads really, really quickly. So it's all about taking the pain from an OpenStack deployment, starting to get the benefits in your business really, really quickly. We do do things a little differently. So here's what really becomes I guess the foundation of a software-defined data center through OpenStack. First of all, SUSE or OpenStack Cloud uses the crowbar deployment framework. Now, what do we do that? Well, it allows us to give a flexible but very, very fast, slick installation and deployment process. Now, just a proof point on that. Those of us that have been at multiple OpenStack summits will recall that Intel used to run a rule-of-stack competition, which was about how to stand up a cloud environment as quickly as possible and then how to build in sophistication, how to make it a true OpenStack Cloud experience, so building in high availability, looking at migrations. They gave time bonuses and time penalties for including more or including less. SUSE won that competition every single summit that it was run. And back at the last summit in Tokyo, we worked with Intel to run a session to tell everybody else how to win the rule-of-stack competition. So quick and easy deployment. We built in the widest hypervisor support possible. So we don't just support KVM for new web-based workload development. We also support Zen, Hyper-V, VMware, and because of customer demand, we've also built in support for ZVM, so building in support for Z-System's mainframe into an OpenStack Cloud environment. Now, why are we doing that? Well, that's because the Mode 1 IT environment, the existing IT environment, can then be brought into the cloud when it's the right time for the business to do it. In addition to all of the new workload developments. We built in high availability. We'll demonstrate that in a second. We were the first to automate the installation of high availability for the control nodes. We've now extended that to high availability for the compute nodes. We, because we have been in open source software for 20 years, we have the widest interoperability for hardware, software workloads to give you a good experience and we can support the whole stack from the guest and the host operating systems and including the OpenStack layer. We also have built in a more business-oriented release cycle. Now, for a lot of our users, RIP and Replace every six months, that doesn't work from a business perspective. We've built in a longer release cycle. We release every year and we build in non-disruptive upgrades for customers. If they need extra functionality from a release that hasn't yet come out from an OpenStack release, we'll work with them to incorporate that functionality. The intention is to build an enterprise-grade OpenStack distribution that works for the end user, fills their needs. Now, I mentioned about the fast and easy deployment and management and I mentioned about the high availability. What we'd like to do is just switch modes slightly here and I'll invite Simon to do the clever and the dangerous bit and do a live demo of those two items of functionality for you. So I'll hand over to Simon. Hi, everyone. Thanks for your time. So I'm Simon Briggs. I work with Mark on the technology stack and I've got a small demonstration sitting on this laptop. You may not have recognized it when you look to the top of the stage but I've actually got enough RAM in here to run a data center. So the virtual technology obviously allows me to run the management platform from SUSE for our OpenStack cloud technology across virtual machines. What I'll do first is move across to my virtual servers so you can see here with the list I've got a series of elements. Okay, so I've got what's called an admin server. So an admin server is our instrumentation platform. It's the platform where the crowbar and Chef technology resides from our packaging tool to then help you build the infrastructure into a cloud in the background. Then I've got actually two control nodes. So one and two listed here and a couple of compute nodes and I've brought them up into a cloud and I've started my controller nodes on the right hand side and they're booting up. So this is an illustration of the working dashboard you get within crowbar and the ability for it to be able to give you instant warnings on a traffic light signal layer as to what's happening and in typical demo style I've just looked across and seen a red light hitting my screen which wasn't meant to happen. So we're having the joys. Obviously this morning we didn't sacrifice enough chickens to the demo gods. So I have been running this for the last four hours no problem and it seems to have bitten me quite hard. So those services are up there. I've got a compute node but from this screen we're able to show you that we're able to take a virtual machine that's empty so here I'm illustrating you plugging in bare metal in your data center. You want to instrument that bare metal into your cloud infrastructure and you just literally turn on that bare metal within the network environment set up for the SUSE Cloud. Of course it's set up to Pixie Boot. You can see here that on my admin node I have a Pixie Boot environment. I have a build environment which then drives down a copy of SUSE Linux Enterprise Server onto those machines. In the background this machine will start building up a small instance of SUSE Linux Enterprise Server and what that allows us to do then is pick it up in our crowbar dashboard and we're able to instrument that node once it becomes available. So I'll just actually reboot the node that says it's bright red. I probably haven't got enough time for it to come back but I can at least do that. So I'll just quickly restart my node in the background. Try and take that horrible red flashing screen. But whilst I've been doing that you'll see that on the left hand side automatically a new MAC address has been signaled so that's my virtual machine that's booted up. Now what do I don't want to do with it then? I want to drive down a personality onto that machine within the cloud. It can either be a compute node so virtualization resource can be a storage node or it could be a networking node. So to do that very simply I'm able to go across it's come back with a red light. That really is fun of the fair. Give me a second we're going to see some hacking in the background sometimes you just feel it's not your day don't you? So if I then take that machine and work on it I can come down here and edit that node. So it's bare metal at the moment. It's got a very small instance of kernel on there that then allows us to register and I can start giving it some personality. So I'm going to call this one Dave. I've got a good friend called Dave and the server finds me of him so I can also there under the role decide whether it's going to be a compute node. So I know it's a compute node. I'm also going to instrument a Docker server. So obviously everyone in the open stack world is very enthused by the power of Docker what it can bring to us. So I'm going to set up Docker on my SUSE Linux Enterprise server to run virtual resources into my cloud. To do that I want to be able to run the booter FS file system. She's just one of the file system supported under SUSE. It's a unique offering of ours. And then I can allocate the node and it's come up giving me an instruction here that has been successfully allocated. Okay, that red light keeps staring at me which isn't very nice but that node has now got a personality and in the background if I show you what's happening that machine is just booted into the standard kernel. Over the next few minutes what's going to happen is my admin server is going to drive down the extra operating system, the full operating system with all the software required from open stack to run it as a compute node within your cloud. This is all reinforcing the fact that SUSE is kind of key methodology process in delivering our solution is to make it as easy as possible to get to the value of the cloud in the background. The crowbar platform is very flexible you can not only use the GUI as is you can also instrument bringing in external drivers from third parties who release some code which is fully supported against the APIs in open stack and you can deliver that very simply because not only is it a standard GUI platform it also gives you the ability to inject code into there so it's a simple case of instrumenting those drivers and you're able to then allocate that resource down onto the machines into the cloud. So that machine now has gone solid yellow it's working in the background this is where I need to learn to tap dance while my controller nodes down it is causing me some headaches whilst that's the case so it's giving me more details about what's happening on that node all the time so as you can see there is a compute node within an open stack environment it needs certain capabilities being pushed down from crowbar as technology plugins and in the background it's just sitting there waiting to go. Let me do a quick attempt at my hack maybe accelerate it it's not playing ball at all okay so on the Susan booth which is number can we flick to B28 I'm afraid I'm giving in there's too many problems. So if you'd like to join us at B28 we'll be more than happy to show you the technology luckily it's running on a series of small Intel boxes the small nukes that you can get and there's a series of boxes networked together so we're very sure that we'll be able to deliver the illustration of the highly available cloud that can show you the power of what SUSE OpenStack is able to bring to you to get a cloud up and working very rapidly and then from a standard platform become extensible as all clouds grow you'd be able to grow through the SUSE OpenStack technology okay so thank you very much for your time sorry about the little hitches that we had I'll go and shoot myself later or Mark's got a gun so that'll be fine cheers