 All right. Perfect. So hello. My name is Sumit Kumar. I work for NetApp. I graduated from Virginia Tech about one year ago. And like I said, I've been working for NetApp as a TME. And what that means is I get to learn about and get involved in cool new projects, travel around, and talk about OpenStack. So how many of you attended the OpenStack Summit in Tokyo? OK, great. So that was my first OpenStack Summit. And I was so impressed by how active and big this community is. So I remember before I started working for NetApp, I was reading up about OpenStack to kind of prepare for my interviews. I remember I thought, well, OK, OpenStack is an open-source alternative to AWS or Azure, and it solves infrastructure as a service. But one more, right? What's so special about it? Now, over the course of this one year, I've come to realize it's capable of so much more, and it can have so many different areas of impact. So today, we're going to go through these capabilities and try to explore the rest of this iceberg called OpenStack. But wait, why should you care? I mean, OpenStack is great at what it does. So why explore these new avenues? Even care to talk about these different domains, right? Now, one of the things I learned at college is that you need to be able to learn, adapt, and evolve. Because if you don't, you're going to fail. And that's not a good thing. So if you think about it, if you're not continuously finding new and improved ways of doing things, you kind of end up like the penguins, right? Now, the penguins are kind of marching towards their own extinction because they're not adapting fast enough to climate change. All right, so coming back to the topic. So yes, OpenStack solves infrastructure as a service. But what more can we do with it? What are some of these other areas of impact? What are some of the other tools that we can integrate with? So let's start off with DevOps because why not, right? Now, there are some industries that can exist for years, even decades, without fundamentally changing how they do things. But we know that's especially not true for the software industry, right? Now, developers are constantly finding new and improved ways of doing things. And that continuous improvement in tools and processes is what led to the rise of DevOps in the first place, right? So the whole idea of a DevOps is you take parts of your development and deployment cycle, and you automate it. But someone still has to come in and write those tool sets. And that led to the rise of DevOps teams. So all of that was working great for a while. But the problem started because these DevOps teams were still managing a lot of infrastructure, like system administrators before them. And that caused problems. So think about it. If I had to, let's say I'm a developer. I want to get a database deployed. Now, those databases still had to be installed, replicated, cached, and clustered. And all of that work was being done manually. And that makes the whole DevOps team a bottleneck. And nobody likes that. And the problem gets amplified because companies and organizations, they want to be able to do more with less. Nobody likes spending money, right? And that kind of puts DevOps teams in a tight situation because now they've got to deal with all these problems and they have limited resources. So how do you fix it? How do you make both sides happier? Now, the one solution is you make it easier for your developers to utilize these infrastructure tools in the cloud so that they can focus on development and not worry about, let's say, backups, redundancy, or uptime. Now, for example, NetApp IT had a similar issue as well. And we solved it by going with the hybrid cloud architecture. And we adopted an automated cloud decision framework. And we made sure that we kept it easy, intuitive, and consistent for our developers to use these tools. OK, so lots of big words, I know. Let me try to explain that visually. So at the top, we have the self-service portal, which serves as a gateway for our developers to come in and request, let's say, storage, compute, or networking resources. So in simple words, think of it as a graphical user interface that allows you to collect the requirements from the developers without making it too complicated, right? Now, our cloud management platform takes over in the second phase when, depending on what kind of workload the developer requested, it decides, hey, which cloud offering is best suited for this workload? And all of this is really great, because the user at the top doesn't necessarily care about what's going on down there. And they don't have to. And we're still able to apply all those best practices without disturbing the user. So next, we have our enterprise, private, and public clouds, all augmented by NetApp storage. And at the bottom, the NetApp data fabric is what allows us to stay in control of our data. It gives us the flexibility that we need to kind of move our, let's say, data around without having to worry about, hey, we've got to pay Amazon or Azure a big sum of money, right? And we all love flexibility, don't we? Nobody likes spending money. So OpenStack played a key role in our solution, and it helped us keep our costs low. It also gave us the control and flexibility that we needed. And the solution helped make our teams happier. And at the same time, we saved a lot of time and money, which is always a good thing. OK, so next, let's talk about Docker. And this one's kind of special for me, because Docker happened to be my initial area of focus when I started working for NetApp. And so how many of you have heard of Docker? I'm going to assume most of you have. OK, how many of you have run the Docker Run command? Or try to, OK, perfect, about putting an application into container and seeing if that works or not. OK, so if you've not heard of Docker before, a very simple way to think of it is think of it as a lightweight alternative to traditional virtual machines that's focused on abstracting away the infrastructure and the associated complexity. So you basically take an application, you put it into container, and now you can run it anywhere. So when I started, again, Docker was my area of focus. I was working for the OpenStack team, but I was working on Docker. So I've been learning about both Docker and OpenStack for about a year, right? And I've heard so many things since. Now, on the Docker side, they're all about immutable infrastructure, right? They want to be done with the complexity that infrastructure brings. And they have a different idea on how community managed projects should be done. And they believe, obviously, that Docker is awesome and amazing is the solution to every problem. But they think that OpenStack is complex. It's about too many things. Now, on the OpenStack side, most of us agree that, yes, Docker is great. But at the same time, we realize that OpenStack is real, that it's helping people solve real problems right now. Plus, we agree that, OK, maybe we can make Docker a part of our ecosystem. We can embrace it. And that's especially true because OpenStack is about way more than just compute. Now, to be honest, I think all of that is great for lunchtime conversations and arguments and debates and whatnot. But I think it's all just noise. Now, the reason I say that is because the infrastructure wars are over and the cloud has won. Be it private or public, real or container, it does not matter. Now, any customer organization that we go to talk to, they want things to be faster and economical, right? And they want the tools that they've already invested in, that their teams already know about. They want these tools to work together, and they want them to work better. Now, Docker needs OpenStack from a manageability perspective, right? Because virtualization on its own is not good enough. You still need to answer a lot of non-essential questions like, where do you put your guests? And when you scale out, how are you going to monitor things? How are you going to manage it? And at the same time, keep things cost-efficient and flexible. And OpenStack can provide a lot of value there. Now, OpenStack 2 can benefit from Docker. Now, if you think about it, under the hood, OpenStack is a bunch of microservices, right? And that can benefit a lot from the flexibility, ease of debugging, orchestration, and speed of Docker. And to be very honest, maybe I'm a bit biased, but I think OpenStack and Docker work better together. Now, the reason I say that is because it's very common to find OpenStack being used in conjunction with, let's say, a private or public cloud. And then Docker comes into the picture, and it provides us with a really simple way to make our applications portable, and it allows us to run them anywhere, right? So think about it for a second. I can now move my workload around and get the advantage, get the best that these different platforms have to offer. And so for example, I can probably use purely OpenStack for my development cycle, when I want to keep my costs low. And then when I go in production, when I need extra compute resources, maybe I can use a mix of, let's say, OpenStack and AWS. So I burst onto the public cloud when I need additional compute resources. And since we're relying on Docker for this kind of portability, you don't have to worry about maintaining different environments or even compatibility between these different cloud platforms. Again, it's a win-win. So it's no wonder that Docker containers continue to remain of great interest to OpenStack user-server respondents. Now, the present approach for integration includes making Docker containers available as a compute node with an OpenStack, and then by adding support for layers further up the stack for, let's say, Kubernetes or MISOs through the Magnum project. I think that's a great first step, but it kind of misses the fundamental values of Docker containers that provide a completely alternative compute storage and networking stack. Going forward, I think we're going to see a bigger push towards a more native integration of Docker into the core of OpenStack. And that, I believe, and I hope, is going to lay the foundation for the next generation of OpenStack, which is going to be simpler and lighter. It's a gold play with Docker. If you haven't already, see what works for you, what doesn't work for you, if not, why not. Now, there are tons of great online resources. We at NetApp are active participants in Docker space as well, so we constantly try to push content that we think is going to be useful to you. So, for example, recently published a blog that walks you through whether or not Docker makes sense for your enterprise use cases. So go check it out and see if that's useful. All right, so big data, next topic. Now, again, this one's kind of interesting because when I was back in school, I heard all sorts of definitions, all sorts of discussions, debates, whatnot, but we all agree that data is important and we agree that data is growing, right? And data is important for your business. Now, if you want to stay competitive, if you want to provide a better customer experience, for example, you want to be able to leverage the data at your disposal. Now, let's take an example. Let's say you have an online marketplace and you want to recommend to your customers what they should buy next. Now, the only way you're going to be able to do that is if you analyze previous trends and previous search results and whatnot. So we want to be able to do all of this, but we want to make sure that it does not impact our cost and performance, right? And then there are other challenges that you need to answer to. So for example, do you deploy a large number of small clusters that, again, comes with its own set of problems? So for example, it adds a lot of manual effort and it has manageability issues, because, for example, again, if I had to get deployed a Hadoop cluster for my big data workload, I'll have to go in, submit a Quest ticket to IT, wait for them to configure everything, deploy, and that adds complexity and time, and I don't like that. So, and again, by taking this approach, we're kind of creating these data silos and it's difficult to move data around and you don't want to do that. Now, the other alternative is you deploy a small number of large clusters, right? And again, it comes with its own set of problems. It's difficult to scale. It's difficult to do multi-tenancy and it's back up in DR, it gets more difficult. Okay, so let's try to solve some of these issues. So how many of you have heard of Project Sahara? Okay, great. So if you've not heard of Project Sahara before, in very simple terms, it's an open stack project which, in a nutshell, allows you to provision a Hadoop or a Spark cluster on top of your open stack cloud. Now, that is great because, one, it's helping you keep things cost-efficient and at the same time, it allows you to be flexible. So, it takes away a lot of those issues that we talked about, right? Since I'm able to now control or edit or make changes to this configuration of my cluster on my own, I don't have to rely on someone else. So I can manage it myself. And since we're relying on open stack, we don't have to worry about multi-tenancy. Now, for the remaining data-related issues, if you use Manila with, let's say, backed by NetApp or SolidFire, you don't have to worry about it either, right? Because we are built for scale and you don't have to worry about backup and disaster recovery and you can freely move your data around. So, you know, it's great. So big data is a big opportunity and that's especially true if you use open stack-based solutions to keep your costs low and stay in control of your data. Okay, so if you're starting to wonder, hey, why is this storage guy talking about all of these different things and domains and areas or whatever and whatnot? So, well, we at NetApp love cool technologies. We like getting involved in interesting, exciting projects. We like finding solutions and, you know, finding ways how we can do something better. And we see data growing everywhere, right? And we think that NetApp and SolidFire can provide you with a smart storage solution. And a smart storage solution can provide you with the flexibility, cost efficiency and performance that you need to do more. Now, for example, if you set up a virtual desktop infrastructure on your open stack cloud and if you leverage NetApp storage, you can provide instant workspaces to your teams and that will make them happier. So make sure you pick a storage partner that's built for scale and performance and can provide you with efficiency and cost effectiveness. So next, I'll have a demo that showcases how NetApp storage performs under pressure. Okay, so this is my performance analytics dashboard. At the top, I have stats for my application layer. And then at the bottom, I have stats for my NetApp layer. So we're going to try and push our workload until we start to see some latency. Okay, so there we go. We see some, you know, 11 milliseconds of latency. Now, this cluster, the storage cluster that I had been using so far was on a hybrid storage platform. So now we're going to go add two all flash fast nodes to my cluster. Now that's done, you're going to see that the average latency will drop back to one millisecond. And the great thing about it is I was able to do all of this without any downtime, right? And that's something you want. Now, if you keep an eye on total number of IOPS and total number of snapshots is you're going to try and push it to the limits. Okay, so it's almost about 2000 snapshots now. So let's say I want to scale out, right? I want to be able to increase my workload. Now it's very easy to do again. You basically add more nodes in a very similar way like we did previously. So it's simple click and push and it's done. Now, if you add more workload to, let's say, your deployment, then that workload gets distributed amongst your cluster. So none of your hardware is getting wasted, right? So the average latency dropped to about 0.9 milliseconds and if you take a look at the total number of IOPS it reached more than a million. And the total number of snapshots and total number of clones has reached a big number as well. All right, so key lessons learned. Make sure you learn, adapt, and evolve because that's very important if you want to stay ahead of the game and stay competitive. Make sure you leverage smart solutions in partners because they increase your potential and they help you reach success. Oh, and most importantly, don't be a penguin. So if you liked what you saw, you can take a test drive. It's free and it's available on our website, netapp.github.io. And please join us tonight. We're going to be at Stack City. We're going to be at the Blackheart Bar and you can enter to win an Oculus Rift, which is super cool. If you want more details on sessions that you're doing related to the topics that I just talked about, I got some trifles over here. You can get them from me or from our booth. We got other giveaways as well. And thank you for listening and have a great week.