 Good morning everyone. My name is Dan Kahn. I'm the Executive Director of the Cloud Native Computing Foundation I would like to particularly thank RPIT and Phil Robb and Jill and the entire ELF events team for putting together this event. It's been incredibly productive three days for me, a really interesting group of folks here that I've been able to engage with and have a ton of meetings. For the last six months, networking has been a major new focus for the Cloud Native Computing Foundation. We're best known for hosting Kubernetes and we have had a lot of success in the cloud and in enterprise software computing, but we see telecoms as the next big market to go after. And so RPIT and I have been literally circling the world together doing a road show in China and then a booth together at Mobile World Congress and such, but it's been a great collaboration. So I think everyone is familiar with Kubernetes at this point. It was originally created by Google but has now had major contributions from Red Hat, from Microsoft, from Huawei, really all the companies, the 17 Platinum members listed at the bottom here and now hundreds of other companies, thousands of other developers. But that's just one of the projects in CNCF. We now have five graduated projects and 16 incubating that are a little less mature but still being used in production by hundreds of companies around the world. So CNCF has now just crossed over 390 members. That's up from 28 when we launched just a little bit more than three years ago. And we're both the largest and one of the fastest growing ever open source foundations. And so I did want to just give a little bit of background on that and then share some lessons that we've learned and approached that might be useful to the open source networking community. Just as a look at that growth, this is looking at Google trends, so what people search for in Google. And you can see that three years ago when we launched, Kubernetes was quite low right with Docker Swarm and Mesa's OpenStack and Cloud Foundry. And other than the fact that people really like to take that week off between Christmas and New Year's, we've had this very steady growth over three years and it continues today. And then on the right side you can see on WeChat search and China has been an enormous market for us. We have a huge number of members and users and certified versions and such and there as well we've just had a huge amount of interest in growth. This is kind of a complicated chart, but it is just a way of looking at the velocity of open source projects. So the size of the bubble is how many unique developers are working on a project, how high up it is, how many commits it's had in the last year, how far to the right shows how many issues and pull requests. And you can see that depending on how you count, Kubernetes is the second or third highest velocity project in open source just behind Linux. And probably over the last three years the program where CNCF has generated the most value and has had the biggest impact on our ecosystem is certified Kubernetes. And this is a way, it's an automated testing system and a process behind that to ensure that Kubernetes that you purchase from different vendors or use in a hosted cloud environment or just download the pure upstream version are all compatible that they all support the same APIs. And so we're really proud of the fact that we have 83 certified Kubernetes partners and we've actually collaborated quite closely with LF networking on the OVP certification program that was announced this week, which in fact is using a lot of the same processes and documents and such. Okay, so that's a canned history of CNCF. What have we learned and what are three ideas that would apply over to open source networking? The first one is a loose coupling. And this is the idea that all of the projects in CNCF do work together. They add value, they solve related problems, but they're not welded to each other. And so different enterprises select different projects that they find useful, different platform providers glue them together or bundle them together in different ways. And I want to emphasize that this leaves space for innovation. It lets people focus on the areas that they're interested in, where their talents are, and it's closely related to another CNCF principle that we're not the king makers. And so explicitly when we pick a project and we say, okay, this is a great container runtime or this is a great service mesh or this is a good tool for service discovery, that is not the end word, the final story for CNCF and what we're endorsing. And we explicitly have a policy that we can host multiple projects in the same box. So for example, we support multiple container runtimes today, and we'll probably see more of that in the future as more projects come into CNCF. So we do think that that loose coupling adds a lot of value and allows innovation to occur at different rates in different parts of our ecosystem. Second idea is to minimize toil. And so we really see developer productivity as one of the most important constraints that we're trying to optimize for. And so that means that we, to the degree possible, provide just a variety of different services to our projects that help make the developers more productive and makes it easier for them to collaborate. In particular, we're perfectly happy to use commercial services to do this when those work well. As an example, I believe Kubernetes now has the largest slack, public slack implementation in the world. It has thousands of new users every week. It's just kind of a crazy process to support. But we also use Zoom video calls and then publish all of those to YouTube, hopefully within a day or so, so that people who couldn't make the call or who need to go back later in review have access to that. We use GitHub. We use open mailing lists. We use a lot of Google Docs. And interestingly, we don't mandate what those tools are to the project. We certainly have some defaults that we recommend, but we're constantly open to trying new ones and to accepting the fact that different projects often want to use different tools and we support them in doing that. Now, I want to particularly call out the role of continuous integration. A key driver of the Kubernetes project's success and robustness has been the level of investment that we've made. Every weekday we run over 10,000 CI jobs. Every time someone opens a pull request, it runs that CI job. And then in addition, every two days we run a new scalability test of 150,000 containers across 5,000 virtual machines. Now, the smaller projects are not at that scale, but they all rely on CI platforms as well. And interestingly, we don't really mandate or even have a preferred one. Different projects use different CI providers, and we're happy to support them on that. And then the third one, I would say, is the value of marketing of vision. And so Kubernetes is the anchor tenant of CNCF and an incredibly important project, but we do try and share a larger vision about what cloud native means about a way of reducing costs and improving resiliency. Most importantly, increasing developer velocity. And so two examples of this is that we don't try and say that the CNCF way or hosting using all CNCF projects is the only way of successfully being a cloud native enterprise, or for this audience, a cloud native telco. We have some recommended options, but we also create this diagram. Can you just raise your hand if you've seen the cloud native landscape before? This has been described to me as useful, overwhelming, and the hellscape. Interestingly, the same person did two of those. It is an overwhelming document. There's a cool interactive tool at the top that you can go to on your phone or your computer. L for landscape.cncf.io that actually lets you research each of these projects and products. You can see that the grade ones are closed source products. The white ones are open source, but we are trying to make a statement here that it's not actually that we're making the world complicated. We're the map maker. The world itself is complicated. There's a lot of different options. And so then the other diagram that we think is incredibly valuable is what we call the cloud native trail map. And this is really designed to be the front side of the document, the landscape on the back. And I will say you can walk over to the CNCF booth after this session and pick up a print copy here. But in my experience, I don't think there's actually an enterprise that is using all 21 of the CNCF graduated and incubating projects and nothing else. No commercial products, no other open source projects. And that in reality, people are mixing and matching to a huge degree. And what we hope is that an enterprise at telco will look at the CNCF project, will evaluate it and try to understand its strengths and weaknesses before making their decision. But there's no obligation to that loose coupling is a really key thought here. So those were the three big ideas. Loose coupling, minimize toil, market division. These are all things that open source networking organizations like LF networking have already begun to do, but could perhaps put more resources into. And then finally, I do want to just mention two other CNCF activities that might be of interest to this community. One is an area that we've been working with LF networking and others on. It's called the cloud native network function test bed. This is a project, it's open source. It's available right now if you'd like to participate where we took some open source network functions out of the own app project and repackage them, not in virtual machines, but in containers, run them on open stack and on top of Kubernetes on identical bare metal servers that have been very generously donated to us by the bare metal hosting company packet. And you can just Google for this CNF test bed and see an online presentation, a lot of detail about it. And then we are having monthly phone calls where we're talking about collaboration. There's some aspects of how we're making this perform today that everything is automated. It's all in the open source repo, but I would describe some of it as being out of band that it is done in a more manual ways. And we're interested in having all of that be automated into Kubernetes. And then finally, I'd be remiss not to mention CNCF's conferences. We host KubeCon Cloud NativeCon. Our event in Barcelona on May 20th is expecting over 10,000 attendees. It's going to be the largest open source developer conference in history that's up from 8,000. You can see our growth here over the last three years from 500 when we got started up to 8,000 in Seattle. We'll then be in Shanghai, June 24th, 26th. That's going to be joint with the open source summit and we'll have a lot of networking content. And then we're expecting an even bigger event in San Diego, November 18th to 21st. And so I hope that many of you will be there and have your colleagues go there. And I'll stop there. Thank you all very much for your time.