 Hello. Good morning, everybody. Welcome to this presentation from Zero Stack. Cool. So we're here to transform your private cloud experience and the next few, next 20 minutes, we'll talk about how we do this. So why has private cloud been so complex, so hard to build? Today, you heard the term super integrators during one of the keynote sessions. Well, fundamentally the problem is that in order to run and manage OpenStack, you need to hire experts which are sometimes quite hard to find and as your scale increases, the number of experts you need increases. The software tends to get very complicated. The picture on the left is from the OpenStack community itself. The picture on the right is drawn by some marketing professionals, but they're both complex, right? On When you when you're running it, the day-to-day, the information that you get out of it does not always provide you enough information to figure out how exactly your cloud is behaving. And the reasons for that is the way people build clouds today is fundamentally more complex. You have to go get hardware, server hardware from somewhere, storage from somewhere, take to piece together multiple software components and then kind of cobble together your private cloud. You're all very familiar with these two sites. You know, the dominant design in cloud was established by public cloud vendors and many people like that simplicity and they want to bring it in in-house in their private clouds because, you know, obviously the compute and all of the data is outside. You get poor performance, sometimes all of that on the private cloud side. Although, you know, customers like the look, the locality of their compute and data, the security compliance, the performance and control, but it gets very complex very soon. What ZeroStack has done is kind of taken the best of both worlds, taken the simplicity of a public cloud and the control and compliance and performance of a private cloud and we've brought it together and in the next 15 minutes I'll show you how exactly we've gone about doing that. We've taken a slightly different approach to building a private cloud. So a ZeroStack is a two-part solution. The first part is the part that lives on your premise and it's built on hyper-converged hardware. You can start very small or you can grow as much as you want on your in your premises and it's hyper-converged hardware. The smallest unit is called a Z-block and it's built on self-healing management software and it uses 100% pure open stack. The second part of ZeroStack is the part that lives on the SaaS. It's a SaaS platform managed by us and that's where you get the cloud management, the monitoring and operations, you know, the self-service, the app store, the capacity planning and charge back. So we're not just talking about, you know, deploying basic open stack. We're talking about a complete private cloud. So who benefits and on what are the benefits that you get from ZeroStack? If you are a user you get, you know, self-service, you get the notion of collaboration, you get much better performance, you get an idea of how much this is costing you, how much capacity you're burning, all of that stuff. If you're an administrator, it gives you significantly reduced operational headaches, right? You get complete automation, complete control over your environment and you get a lot of insight into, you know, how to plan your data center, you know, what the cost is and all that stuff. So the main thing is that it's not only open stack, it's a complete private cloud. There are no complex software integrations. So what that means is you can spend time on your application and not so much on your infrastructure because your business runs on the applications that you run, not on the infrastructure, right? And you can start very small and you can grow as your IT needs grow. Let's drill a little bit into the specifics of what you gain. If you're an administrator, you get the cloud dashboard, you get monitoring and analytics, you get a global infrastructure view. This zero stack platform is multi-site aware from the ground up, okay? You get a complete capacity planner, you get usage and charge back information. It's integrated with your enterprise with AD, LDAP integration. For users, you get self-service, you get an app store, whether you're deploying a single VM app or an app that's a multi-VM app, you can deploy all of those. It is multi-site enabled so you could deploy an application across different availability zones and regions and you get the notion of a project planner. It's all API and CLI enabled and you have role-based access control. Now, what are the things that you don't get because of zero stack? And we call it the zero stack unexperience, things that you don't have to deal with anymore. There are no special controller hardware, so we call them management puppies. There are no management puppies in your data center with zero stack, right? We take care of that problem. Zero stack is self-healing and what that means is it's HA by default. You do not have to decide and pre-configure HA. It's always HA. It keeps the clouds always on and the system is based on predictive analytics. It predicts issues before they actually happen. And in terms of patching and updates, that's very iOS-like. It's handled by the software itself. So as you can see, there's no need to configure, troubleshoot, all the disparate software and hardware components to kind of build your cloud. In terms of what it's powered by, it's built on 100% pure open stack. All of the APIs are 100% pure open stack. We are currently on the Kilo release. The services that are enabled are out here. You know, the Keystone Nova, Cinder Glance, and all of the Devcore services, the Heat and the Object Store. So without further ado, let's do a little demo of the product. This is a recording of a real demo. So first, we'll go through the process of how easy it is to create a cloud. So the setting here is that we've shipped a single Zblock to a customer. They've racked and stacked it, and they can ping Google.com. That's the only requirement. Then everything else is done from the SAS portal. You register your account, and you get a unique account ID. And you use this account ID and associate with the four hosts that make up a single Zblock. Then you log in to the system, and there you see all the four hosts. Now you provide the east-west networking traffic. We provide the four different tunnels for your clustering software, the management software, the VM traffic, the storage, the scale-out storage traffic. So out of the box, we provide four different pools of storage. These show up as volume types when you're creating VMs. So you can select and choose all of the four pool types. On the physical side, we're organized into availability zones and then regions. A number of hosts make an availability zone. A number of AZs make a region. You see the environment one. It's all connected to Active Directory, or you could create local users in Keystone. You get a summary of the environment of the cloud that is being created. You know, you get an idea of how much capacity you get out of this cloud. And you say, build my cloud. And this takes anywhere from 10 to 20 minutes, depending on the size of your infrastructure. Notice that at the end of this, you'll get a complete private cloud. I did not say anything about Chef, Puppet, Python, Juju, Ansible, nothing. We believe at ZeroStack that to build a private cloud, you should be able to do it with ease. Those orchestration tools are meant for your application layer. So now your cloud is ready. And so now let's go through. Once your cloud is ready, you can connect as a cloud admin. And you log in and you kind of get a cloud dashboard, all the stuff. On the infrastructure side, you see your regions. You see your availability zones. You can add to your availability zone. You see all the four hosts. You get an app store. You see the apps and images. Right. You can configure it in whichever way you want. You want to create an extra, extra large VM flavor type. You can go ahead and do that. You configure your external network. You kind of do all of that. So that's what you get. And then let's go through the process of now that you've created a cloud, let's start by creating a business unit where your end users can actually start consuming that cloud. So it's very simple. You could decide where to connect that business unit to whether to an active directory or just for now local. We created a business unit called Corp. And it's that simple. So now I log in as the admin for that Corp. And now I can go to the business of creating my projects. I just create a project. And I can customize that project based on the needs. I can set all the quota and all of that stuff. I can add as many users from the business unit, from my ADL app, all those things. I create a private network, create as many security groups, all that stuff. I can add some key pairs to it. And here I can decide if my project is going to be short-lived or long-lived. A lot of times when you create a project, you don't know, you know, some people take resources and they don't want to give it back. So you can, as an admin, set these ground rules. So you're in complete control of your situation. Right? You can associate it with the cost center to kind of initiate, you know, the notion of showback or chargeback. Right? And there you go. You can have your project. I could take another, create another user and make him an admin in that project. I could do that in multiple places. Right? So here, if you notice that everything that... Okay, let's go into the details. Right? We're creating an image, doing all of that stuff. You can configure the networking. Right now we have two private networks. So let's create a router. Kind of create a link between them. And you kind of see that now your private network would be kind of connected to your external network. We've just established a link between the two. Right? Let's go ahead and create a VM. You can decide whether you want to create a VM from an image, from another VM, from a volume, from a snapshot. All of the ground rules that were set in place by the administrator kind of get applied here. You can choose the type of volume. You can assign a floating IP at the beginning itself. So notice that this VM's got created and at the bottom of the VM you see the timelines. So everything that we do in the ZeroStack world that all of that information is saved in a timeline view for you to come back. So that's where it helps for in troubleshooting, for auditing, all of these capabilities. Whether a VM has been snapshotted or imaged or a VM was powered off, whether VM's performance number is shot up, you can see all of that in a timeline view. So you would come back in time and kind of see a history of that particular VM or for a project or for a host, all of those things. So there you see the VM has got the private network of the network that it was attached. An IP address from the private network that it was attached. It's also got a floating IP and all of that. So you can see all of the metrics and you offer that particular VM by select. This is a very new VM so there is not much useful information but if you were to select, you could select all of the metrics of that VM and the data would show up at the same time. So let's say there was a bad event, you'd see a red line going through all of it and you'd see the values for all the metrics and that is very useful to somebody who's trying to debug what is going on with the VM and you could juxtapose it with host information as well. Now this is a pretty long video so I'll actually shift over to the last one that I wanted to show which is a demo of, I don't have much time, so which is a demo of the high availability feature of the system. So this, pardon me, this got a little bit of CLI in it so I'll just show from the beginning, I don't want to show all of it for the lack of time. So we're shutting down a host, all right? And so this in a single Z block, even in a single Z block which is a four host system, it is highly available. I just want to show how it is that we're shutting down a host. Let's try to log in. And if I log in now, I can still log into the system despite a host having been shut down. I go to the host and I see as one host is red. So that's the host that is not reachable, it's disconnected because it's been shut down, right? Now let me try to log in as the business unit administrator. I try to log in and I'll try to create a VM out of the environment, right? I can still go see all of my workloads and I can create, oh, this is very small. It takes a little bit of time to create the VM. So as you can see, we successfully created a VM. So the system healed itself and as a customer, you really don't feel anything. You did not have to configure any, you did not have to configure any HA separately. There was no need to set up any separate controller nodes. So this is what we think is a great approach to building a complete private cloud. It's 100% open stack but with a lot fewer of the hassles. Thank you very much.