Cloud Native Computing Foundation KubeCon + CloudNativeCon Europe 2018 (Copenhagen), 14 Nov 2015

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Keynote: CNCF 20-20 Vision - Alexis Richardson, Founder & CEO, Weaveworks

Description

Want to view more sessions and keep the conversations going? Join us for KubeCon + CloudNativeCon North America in Seattle, December 11 - 13, 2018 (http://bit.ly/KCCNCNA18) or in Shanghai, November 14-15 (http://bit.ly/kccncchina18)

Keynote: CNCF 20-20 Vision - Alexis Richardson, Founder & CEO, Weaveworks

About Alexis
Alexis is the co-founder and CEO of Weaveworks. He is also the chairman of the TOC for CNCF, and the co-founder of the Coed:Code meet ups. Previously he was at Pivotal, as head of products for Spring, RabbitMQ, Redis, Apache Tomcat and vFabric. Alexis was responsible for resetting the product direction of Spring and transitioning the vFabric business from VMware. Alexis co-founded RabbitMQ, and was CEO of the Rabbit company acquired by VMware in 2010, where he worked on numerous cloud platforms. Rumours persist that he co-founded several other software companies including Cohesive Networks, after a career as a prop trader in fixed income derivatives, and a misspent youth studying and teaching mathematical logic.
Join us for KubeCon + CloudNativeCon in Barcelona May 20 - 23, Shanghai June 24 - 26, and San Diego November 18 - 21! Learn more at https://kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy and all of the other CNCF-hosted projects.

A

Hi there, everybody Thank You Kelsey that was really cool, okay, so I'm going to talk about the vision for the CNC, f and, of course, naturally, because we are a community. This is very much my personal take. So please don't go up afterwards and say to Dan. Why do you think these things that Alexis said? Because you know we're all different and we're converging over time, but first welcome to Copenhagen, which is the Harbor City, the right fitting place for kubernetes and docker notice, the English spelling there for the VMware guy? Who can't spell harbour.

A

That's Hans Christian Andersen by the way played by Danny Kaye, and it's also the home of Lego, which is my theme for this morning. Who's used Lego before this is the NASA commemoration of women. Scientists and Lego means play well. You can do amazing things with Lego, it's a real source of sorry Kelsi innovation. Here we are innovating by making a tiny Big Ben. Here we are making a real car and octopus eats a spaceship. That's pretty cool I, don't know what that is, but it's definitely crazy.

A

So 2016 was when things started to move forward for the CNC F and it pretty much looked like what you see here, a heap of bricks on the table and we've built something from those bricks and I can now show you the plant at last is here's what we had in 2016. We had a really great tool: Cuban Ares, a few other things in the pipeline, and we had one person who was really good at talking about it as Kelsey there. That was his original gravitas, thanks Kelsey for taking us through that phase.

A

So where are we now I? Think the key point is the startup phase is over think of CNC F in those first couple of years, as being like a startup, a group of people with an aspiration and a hope and a vision to build something, bigger and I. Think after that phase is when you first start to see what you can build and that's what I'll talk a bit about today is building the next thing.

A

Here's a lego house. What we're building today is acceptable solutions for real companies, and we just saw an incredible example this morning, so not everyone's like certain. It's also for small companies as well small teams, but anyone can build things by putting the components together, I really liked it. When we had to mention earlier of oh I, just did this with Prometheus and Griffin or and kubernetes I thought. That was a great example of what we're all about here.

A

But what comes after that in the next few years, is beyond acceptance, wider adoption, we want ubiquity, you want everybody to use this and be super excited about it and do their own thing build their own Legos on top. So what might that look like, so we're we're not a start-up anymore. We're accepted we're growing. Now, we've seen some great numbers this morning. The pieces just work together. What are we building we're building a cloud platform so I'll talk about what is a platform?

A

Why we're building it today, we as everybody in this room and Beyond the whole community. So who is it for? Why do we care, ultimately it's the end user and that often means business users, and that means we have to accept that for many of these people, cloud native is just cloud. Okay, it's very exciting for many of us who are deep into the technology to talk about cloud native and serverless and powers and these things, but for these people it's cloud and they want something simple, a cloud platform for the era of Ubiquiti services.

A

They want to roll out over the next few years. This is a bigger deal in the internet appearing in the 90s and it's open like Linux. So the analogy here is an explosion of companies, projects, people doing things just like the 1990s, but much much larger across the whole world. Based on a single open platform, it : to Linux today and for the business folks, what they're looking for is outcomes like new business models?

A

So what are these business models? Here's a photograph I took about five years ago from a big bank, consulting project, doing Hadoop and, as you can see, there are a lot of consultants involved, and this is all to do. Big data mining or something I, don't know. If we want to do that today. What do we do? It's super simple, okay, I mean it's not quite this simple, but it's pretty damn easy, I go use. Cube flow I can install that in five seconds on cribben eTI's.

A

Now, admittedly, we're not at the state of Nirvana when it comes to installing kubernetes, but I can go use gke or e KS or aks and I can then use cube flow, which means I can be doing machine learning in the time it took for the business person to return from the bathroom. This is great. This is so much better than what we had before. So this is a real business change and this is velocity we're looking for on top of this new platform, so underlying this is something called componentization which I've talked about before.

A

Componentization is all about how you can source multiple components instantly from anywhere to build things very fast, and that collectively makes a platform, and this isn't unique to technology. Here's an example from China. This is a real example.

A

This is a photograph from the Guardian newspaper last year, a Chinese construction firm built a skyscraper in under a month because components they can get whatever they need from anywhere and if something doesn't turn up or breaks, it can be replaced by something else very quickly, because things just work together, interoperability, the basis of any platform as I, think sample of what happens. If you don't do this here is a book which I highly recommend that you look at.

A

If you get a chance by a gentleman called Thomas Thwaites who decided to build his own toaster from scratch on the front, you can see the toaster that he built I'll. Tell you more about that in a second. He did not use any sourced components. Everything was built from scratch by hand, including you know, the wire that connected the toaster to the electricity.

A

So he had to build the plug the cable, the bits inside the toast of the bit outside the toaster or from scratches we come over a year and then he turned it on and it fritzed immediately, which is why it's blown up on the cover of the book. So we don't want to do this. You know we want to go quicker. Whoops hang on, we go back yeah. We want to go quicker. We want to achieve what we can see if we read about velocity in delivery. This is from the puppet labs survey of 2016.

A

You can see these numbers again and again and again, there's a book called accelerate by Nicole for segrin Jeanne, Kim and Jess. Humble I recommend talking about this phenomena of how to go faster by deploying more quickly using automation and tools like Cuban Eddie's, and the key point is when you move from one change a week to many changes a day. You get a lot more confidence in how you work, which is a phenomenal uplift to the confidence of the team, and you can start to do things that you didn't dream of before.

A

So it empowers you to innovate so you're out of that world of homemade toasters. So the outcome we're looking for with this cloud platform, is higher speed and lower barriers to entry so that everybody can do it, which will then lead to an explosion of what people call higher order systems things on top of the platform that are cool. In other words- and you know you have a platform when a lot of people do this at once, and you see this explosion of innovation. Sorry, Kelsey I'll apologize every time, I say innovation Kelsey.

A

So what are the ingredients of this developers very important? You can't have a platform unless you have developers, you know how to use it. They're going to write something called code code is what powers applications. So we need tools for that making applications. We need a marketplace so that we can share the results of each other's work and we need a place to run it, which needs to be easy and stable and operable, and we need best practices.

A

So here's a rough idea of what that might look like using my appalling graphics and the key point about this diagram. Is this number one? It runs anywhere the stuff that we're talking about the cloud platform runs anywhere: Amazon Google Microsoft, your own data center, maybe edge computing drones.

A

Who knows there is a layer of Technology, which is what we're building out at the moment, which is necessary to run the code on top and that's provided by the anchor2 component kubernetes, the other CN CF projects and, of course, each individual place that you run your code will have local services and data like on Amazon. You might use sqs, you might use something else, pub/sub on Google Cloud and the key point is if you're, an developer who's going to build stuff to make the business happy.

A

You want a way to say to what what you want a way to say. Please just run my code, so instead of thinking about deploying containers, you need the ability to say run my code. Okay, this is the key change and I think this is what's still missing from the world of CNC F.

A

In terms of this explosion of innovation and in terms of the higher order systems that I mentioned and things you can build exciting Lego projects you can build. We are seeing a lot of projects appearing at the moment, and this is what I think is evidence that the cloud platform is the right one and can be built in the next cycle. Serverless, for example, there are lots and lots of great projects implementing different concepts around serverless platform as a service, then you have things like machine learning, service mesh.

A

There are interesting projects for doing declarative, apps like ballerina caisson it composed from docker and so on. There's things like laravel, which takes PHP and makes it kubernetes ready. This is all proving the point that this is innovations already happening, there's much much more to come, so the platform will run these things. So there's an opportunity for me to say a few words about serverless where we are heading to I.

A

Think is these concepts are serving us in even at ease and cloud and cloud native, but we'll stop being seen as different in in not not too short a time and developer will get to the point where they can say.

A

Kubernetes just run my code instead of running my container and we're starting to see tooling appearing like meta particle from Brendan at Microsoft, which is general-purpose but also use case, specific tools that makes consumption and packages of service of server, less style just run my code very easy and what's going to drive, this is the ability of kubernetes. So there's that point about obliquity again and so by 2020.

A

We will see new programming tools that unify these concepts, and these will be cloud agnostic, multi cloud, if you prefer, and the business models that people talk about with server lists, will fade into the background. They'll be part of the management platform. So this is the convergence.

A

So in terms of a sort of very back-of-the-envelope technical roadmap from the TOC, it might look something like this. You know on the left. You can see what we've done, which is about putting in place the core foundations, the core platform, the engine room, the orchestration, the containers, the networking, observability and operability next and then routing or if you prefer, routing mesh messaging, G RPC things like that, but this year it's about security, there's been so much from the CNC F on this, but also projects like spiffy, inspire, OPA storage is a really big deal.

A

We're not done until we have storage fed into the platform and the stuff happening this week on that interfaces. So open metrics is spun out of Prometheus it's taking the model and standardizing it so that anyone can do it for other projects as well. Open events came out of service. This is a way of doing passing events in a functional style and the finally the developer on-ramp.

A

This is something else that we must have for the platform, and once we have these things in place, then we need the ability to have the add-ons that you can just say just run my code, so that is where we are going so a few other things practice when you have a new platform and new tools, you have the single curve, Ellucian practices- and this is often very emotional, a very tribal.

A

So you have this rise of DevOps in the area of cloud now, we've got a new tribe of people, you who are going to be doing things with the cloud native cloud platform I've been going around calling this get ops. Other people have been talking about it too. Pushing code, not containers, do not doing operations by pull request, empowering developers with operational control through making get the center of control and here's a picture of that.

A

The key point is the developer. Experiences get push, which means you can be completely code centric and this model works not just for kubernetes, but also for anything sitting on top of it and that's the beautiful part, and it gives you continuous delivery, which gives you velocity, and you know that it's safe, because every update is atomic. That is really important.

A

So to summarize, how do we go faster? We have a cloud platform powered by kubernetes and the CN CF tools. It is supported by everybody. You see an explosion of new forms on top like cube flow like sto, like pachyderm, like laravel, like open fats like so much more cube less.

A

You have a high velocity model for doing practice of get ops, and that means that kubernetes ubiquity is kind of a virtuous cycle and yet the point where we'll stop talking about using software unless it is in some way associated with kubernetes for these use cases.

A

So what's missing anything the developers. Of course it's up to you to make this happen and build things out of the Legos. There's going to be a lot more of you in the future, about 100 million I. Think in 2027 from the current low 20 millions.

A

So what will you be doing? Oh bit quitter services and we're heading to a future where I know if I'm happy about it or not technology being part of everything? Isn't it scary to think that I could go to somebody's house and you would know who I am, and it would give me access to something that terrifies me. It's possibly a dystopia, not necessarily a good thing. So that means there's something else. I want to finish on. I think is very important and that's ethics.

A

With this enormous power that we're being given with these tools, you have to use them responsibly. It's that's up to you, it's not something that other people will do for you, things like diversity. You know I fully agree with the comments made by Kelsey and the previous speaker on that point. But this is table stakes. We also have to act ethically and morally, and that means we have to lower the barriers to acting morally, which means that we have to invite more diverse moral people into our community.

A

It cannot be a set of people who are all the same. Who may therefore make unchallenged assumptions. This is a key point, so I hope you think about that today. So finally, don't be this guy.

A

Be these people? These are the future.

A

Thank you very much.