 Hi, good morning, everyone. And I'd like to thank Angela and Katie and everybody on the team here for putting together another fantastic event. I hope everybody had fun in getting to actually go to a castle last night. And I enjoyed when I Googled it, seeing the pictures of Angela Merkel and Barack Obama having dinner there just, I think, a couple months ago. So if we could just have a quick round of applause for them for putting together what I think has been a really well-organized. So I just, we'll try and keep this relatively brief, but I did want to give you kind of my history of the cloud and how I perceive this. Before I start, I'll just give a quick overview of the Cloud Native Computing Foundation. We're actually less than a year old. We're one of the very newest projects of the Linux Foundation. We're best known for Kubernetes, which is the orchestration and scheduling platform for containers. It was donated by Google to CNCF, and Google remains the biggest contributor to it, but we now have tons of community support, also from a lot of big companies and committers from a lot of other companies. And I'm pleased to say we've also now added in our second project, Prometheus, which actually got started right here in Berlin. And if you do have a few minutes later today, I would encourage you to go by the CNCF booth where we have several of the Prometheus core committers, and they can show you a demo of that and some of the other work. This is a really powerful platform for watching your distributed system and measuring it and tracking how things are going. And it seems like it's becoming something of a standard for many different components to interoperate with. So CNCF has its own board. We have these Platinum members listed here. We are actually up to another 50 members, and we're growing fast. Over the next six months, hopefully you're gonna see us add in another half dozen or more projects. And what we're trying to do is build a set of open source stacks that allow people to build a complete cloud solution that can work on multiple different clouds and across clouds, and I'll talk more about that. But first, let's just go back in history a little bit to the year 2000, where when you wanted to build a new application, in most cases, the building block was a physical server. You needed to call up your sun salesman and say, please send me one, and it would often take weeks or months to get one. Sometimes if it was a big application, you would buy a rack of physical servers. My first startup was an internet e-commerce company way back in 1994 that set up the first music store on the web. We used deck Ultrix boxes that we had under our desks in a house in New Hampshire, but it was the same idea. When you needed more capacity, you just added on a new box. So this worked great, especially for sun, especially in 2000, they were selling loads of these things, and so then our next stop in our history tour is VMware, and they had the relatively brilliant idea of virtualizing that server so that each application could have its own virtual environment, and you could have multiple different applications sharing the same physical server. Now, and this is really based on the not so well kept secret that the vast majority of servers in the world are running at well less than 15% utilization. I've heard statistics from some enterprises of 8% or even 3%. And so the idea is that if you can virtualize it, you can pack a bunch of different VMs onto the same machine, and now when you want to run a new application, you don't need to buy that new server, you can just spin up a new VM. So the architectural building block became a VM, although it was still relatively heavy weight, had its own complete operating system unique for that application. So now we jump forward to 2006, and my next speaker will be Chris Schlager from AWS, who really, you know, AWS totally revolutionized the world here by saying, and I'll say by popularizing, there were a few others before it, this concept of IaaS that you don't need to buy your servers anymore, you can rent them by the hour. So this capital expense, when I was a venture capitalist, we used to say, oh, it was, you know, a couple million dollars to buy all the servers, now it's just a few thousand dollars a month to rent them. You don't need to buy the capacity until your demand is actually there. But the building block of what it meant as you were building each part of your application didn't change that much, it was still a virtual machine, it just changed its name from a VM to an AMI, but it's still basically the same thing. And then the last stop on the proprietary side of our tour is a company called Heroku in 2009 that had a huge impact, I think in helping people understand how much better developer processes could be. And the idea here was that if you followed a set of relatively clear rules, called building a 12 factor application, that you could deploy your application into their PaaS without having to think of all of the details about operating systems and versioning and keeping things up to date and you didn't necessarily need to hire the Ops staff and the Dev Ops staff. And there is this magic when you're first learning it and you say, oh, okay, I've made these changes to my app and you can just type git push Heroku and about 60 seconds later, that new version of your application is up live on the web. And so I think, and Heroku really had a massive impact on the industry about thinking about trying to abstract away a lot of those details. Now, often as with guide rails and other sorts of areas, it can constrain more than people want it to, but for getting started, it's absolutely fantastic. Okay, so now we go to our next step in our history, 2010. And interestingly, the next four stops are all in the open source world. So you had this consortium open sac startup, originally Rackspace and NASA, but it now includes most big computer companies. And they said, could we go compete with AWS and VMware? Could we create an alternative to those platforms that's completely open source that anyone can make use of? And I think looking backward, we could say, oh, okay, maybe this took a little longer for them to build than they thought it would. Maybe it was a little clunkier in different ways or a little harder to install. But at the end of the day, they were successful. They have a platform that people can install and work and does create a viable IaaS competitor. That is, it offers the sort of EC2 VM competitor. It allows you to do a virtualization. But the key thought is that the building block of what OpenStack offers is still very much based around a VM. So then the next year, Pivotal provided a service called Cloud Foundry that they open sourced several years later, and that foundation actually lives next to ours in the Linux foundation that have an open source project that anyone can use and contribute to. And that's been relatively successful in saying that this Paz concept, this 12-factor app, you don't have to just use Heroku. There's now a number of other platforms that basically give you the same capability and mainly offered by other companies. Okay, so now we jump to 2013 and we had Solomon Hikes here two days ago from Docker. And this technology really upended a lot of the previous ones. And what's interesting is that there's no single part of Docker that was entirely new. He took different technologies that were actually already out there being used in different ways. But sometimes just combining things together and making a real user interface around it and marketing it actually does change the world. And there's an analogy I used to Tim Berners-Lee in the World Wide Web, where if you look at the components of HTTP is basically like FTP and HTML is kind of like XML and URIs aren't actually all that new. And yet somehow the magic of those three together created the web. And so similarly, Docker has a really revolutionized software development. I think the vast majority of you are either already using it or considering it, but this idea that you can create an exact development environment. And in my last startup, we used to spend so much time and have this long instruction manual for a new developer coming on board and they would get their Mac or their Linux machine and they would follow through all of these steps of the package manager of installing everything. And at the end they would hit some bug and there'd be some command line that or some dependency that had changed from when the instructions had been updated a few weeks ago. And so this concept of immutability of being able to create an environment and then being able to reliably share it between machines. And so I'd say that Docker has had the fastest uptake of a developer technology ever, but has only in the last, but more recently have folks started figuring out how to run those Docker containers in production. And so that brings us up to the present of what we call cloud native technology. And we define that as having three key components to it. So you have microservices where you divide your application up into multiple different parts. You package each of those parts into a container and then you use a tool like Kubernetes to dynamically orchestrate those containers in order to optimize your resource utilization. And that's basically to save costs by using the smallest number of machines possible that can fully meet your requirements. And part of the magic of that microservices is that you no longer need to be locked into one operating system or programming language or platform for your application, but you can divide it up and have optimized each part based on its requirements or what that part of your development team wants to use and have those different parts communicate each other with each other. Okay, so now let's sort of look back on that tour and say what have we learned, what's been the progress? So the core building block of applications has gone from servers to virtual machines to build packs to containers. The isolation unit went from heavier to lighter weight. The container shares the operating system between the different versions of it. That's both in spin up time and in size. There's this concept of immutability that instead of each server being totally custom that you're configuring it just for that use that it's more like cattle that it's mass produced and dealt with. And then the provider has gone from closed source single vendor to open source available for many different vendors or that you can just download and work with yourself. And just a quick mention of Paz because we do think that there's just a ton that people have gained from Paz. But what's interesting is that several CNCF members are actually building Paz's on top of Kubernetes and other cloud native orchestration platforms. And so the idea is that many applications actually start out as 12 factor, but then sometimes they outgrow that paradigm. And so what's nice is the ability to have those guide rails when you need them, but then when you need more power to be able to break out of that and to grow into a slightly more complex model. And so Paz on top of cloud native supports both. Okay, so what's the advantage of this? The first one I wanna emphasize is isolation. This is the familiar Docker whale logo talking about different whales of the world. And the sort of magic of Docker is that dev prod parody. You get your entire environment working in development, you freeze that image and you can now copy it and push it up to production and move it around and replicate it and you're getting exactly the same thing. And that fosters code and component reuse and it can just dramatically simplify operations. Next one is lock in that the value of open source software, which is I think one of the biggest messages of LinuxCon and ContainerCon is that it means that you're no longer locked into a single vendor. So there's multiple different vendors that can support you that you are no longer locked into a single public cloud or private cloud. You can use the one you want to, you can move between them. And also interestingly now you can use them in combination. And so that's a functionality that folks are increasingly looking at. Unlimited scalability. So Google has been using container infrastructure for over 10 years. They have a statistic from two years ago that they launched 2 billion containers per week. That's 3,300 per second on average. But of course their peak is much, much higher than that. So the idea is that these platforms are optimized for tens of thousands of self-healing, multi-tenant nodes. And this slide is talking about the value of microservices. I'll just say the ideas you have these multiple containers and they're communicating to each other. But boy do I have a lot of respect for these pirates for how much effort it must have been to get these containers onto those rickety little ships. But the idea here is that you divide your application up and they can all communicate. And those different containers, maybe you start out with a Rails application and then you cut out the credit card piece and that can be a Python one and then something that's more performant you can rewrite and go. And so over time your application can be cut up into different pieces and optimized for that part of your team and for that function. The efficiency and resource utilization is kind of the basis of orchestration. So the idea is that you have these different containers that you're defining what the requirements are and your system is dynamically responding to those requirements and to the needs and the demand auto-scaling. And that both reduces your costs of maintenance and operations. It also just reduces your server costs. And resiliency. And so this is a relatively shocking image of a container ship. Thankfully, no one was killed on this. But individual containers, machines, even data centers can fail and you can configure your platform, your cloud infrastructure to be able to survive that. So that's the kind of basis for what the cloud, our vision for cloud native. The idea for CNCF is that we are not trying to write that software ourselves. We're actually trying to be one of the best platforms for hosting projects out there. And so many of you are probably involved in different open source projects. We are interested in hosting those and we're talking to a number of different projects. I'll show you in a second. Our view is that with GitHub out there, the role of software foundations has changed. That us just providing a free mailing list or a website or a repo is no longer that interesting to people. That we need to provide a new set of services. And our goal is to be the best placed host cloud native software projects. So the reason that we think that projects should come to us, the first bullet point on there is by far the most important, which is that a neutral home for a software project increases contributions. So a project like Prometheus that actually started in SoundCloud, but now has committers from CoreOS and Google and Weaveworks and elsewhere. So that neutral home allows more people or makes more people comfortable contributing. We have a really a gust technical oversight committee of top architects, folks like Solomon Hikes from Docker, Ben Hindman from Mesosphere, Brian Grant from Google and others. And there's an endorsement, essentially they act as gatekeepers. And I do wanna emphasize that it's that technical group that accepts in new projects, not our governing board that's made up of vendors. We have a $15 million thousand node cluster of city yard Xeon servers that is available. And I will mention this is actually available to anyone for your own projects that you can apply on our website and make use of it. But we provide priority access to the projects that we host. We have an end user board. We have a press relation and analyst relation teams. We provide some cash to improve documentation, which is always a weakness for every project. We don't try and rewrite projects governance process. We just ask them to have one and to have a fair process for dealing with it. We have a full time staff starting with me that's happy to help them. And we have a really great events team, which is actually the same group that put on ContainerCon and LinuxCon here and are now helping us put together instead of we have a worldwide meetup groups. We have a set of road shows that we're gonna be doing around the world to kind of spread and evangelize this idea of cloud native. And we have a marketing demo where we're showing how all of these different technologies can work together. So I guess I will encourage you to take a picture of this screen if you want to, just to be able to prove six or 12 months from now, how bad my prediction capabilities are because the only thing that I can guarantee you is that all of these projects will not be part of CNCF a year from now. Some of them are technical oversight committee will not approve, some of them will change their mind and not be interested in pursuing it with them. Some of them sort of better alternatives will come along or there'll be a merger or some sort of competition. But I think that this list is still useful because it gives something of a concept of the kinds of projects that we're looking at and the kinds of spaces that we're exploring. We're not trying to do our own Linux distribution, we're not trying to go super low level to firmware or anything, but we are interested in a cloud database, for example, CockroachDB, which is something of an open source implementation of Google Spanner. Messaging NATS is a neat kind of competitor, sort of more modern competitor for RabbitMQ and a lot of other ones. And I'll even give you the heads up that I think Open Tracing next week is probably gonna be the third project that gets incorporated into CNCF and that we're very excited to have that happen. So this is a little bit of a view into our future. Our goal is that we do have a number of new projects that kind of fit together. There's no expectation that end users need to use all of these projects. The idea is that we provide an open source stack that users can make use of, that we know that these work well together but that users are also welcome to just pull out specific pieces of or replace any given component with their own version or proprietary one or a competitive project. And finally, I would just ask all of you to please consider getting involved. So our projects are always in need of new committers, bug fixes, new users. We have a set of meetups around the world that you can go to our website and see and learn more about cloud native, Kubernetes, Prometheus, the other ones. We're gonna be doing these road shows. They're gonna be targeted to a slightly less technical audience than this one but maybe for some of your colleagues who are just getting started in this space but our goal is to actually have those go around the world. I certainly would encourage you as you're looking at the next versions of applications that you're deploying in your companies to be thinking about a cloud native platform of orchestrated containers of microservices. I think for many legacy applications, running them inside of a VM makes a lot more sense and maintaining that is a much easier thing to do but for that next version of it or when you do update it, moving to a containerized microservices platform is really the future. And then I certainly would ask you to consider joining CNCF both as a vendor and we have next week we're gonna be announcing a much lower cost end user membership as a way of participating and tracking our progress. And then finally, I just wanna give a call out. Our big event, in fact, our first big event is in Seattle on November 8th and 9th and this has an even longer name than this event here, cloud native con, cube con, Prometheus day and we're expecting to sell out at about 1,000 people. So if you're interested, I definitely would encourage you to take a look and we have the pretty amazing list of speakers on there. And then we will be in Europe in April of 2017 and we'll hopefully have some details on that in the next few weeks. And so if you can't make it over to America, we'd definitely love to see you at our event back in Europe and before in another six months. And please feel free to reach out to me. I'm very happy, I'll also be here afterwards and I've really enjoyed the hallway track of this event but please feel free to email me or reach me on Twitter and we have a good amount of information on our website to find out more about the project. So thank you all very much.