Cloud Native Computing Foundation KubeCon + CloudNativeCon Europe 2018 (Copenhagen), 4 May 2018

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Keynote: Accelerating Kubernetes Native Applications - Brandon Philips, CTO of CoreOS, Red Hat

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: Accelerating Kubernetes Native Applications - Brandon Philips, CTO of CoreOS, Red Hat

CoreOS introduced the Operator concept to the Kubernetes community in 2016: a tool that automates the operational tasks necessary to deploy, update, and run a distributed application on Kubernetes. Operators reduce the barriers to entry to develop applications on Kubernetes - but building an Operator from scratch is a time-consuming task that requires a deep understanding of Kubernetes. Join Brandon Philips for a look at what’s new to help accelerate your ability to build your own Operator and get back to developing your innovative application.

About Brandon
Brandon Philips is helping to shape the future of modern Linux server infrastructure at Red Hat. Brandon joined Red Hat as member of technical staff following the acquisition of CoreOS where he served as CTO. Prior to CoreOS, he worked at Rackspace hacking on cloud monitoring and was a Linux kernel developer at SUSE. As a graduate of Oregon State's Open Source Lab he is passionate about open source technologies.
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

Good morning, how is everybody you've been very quiet, very, very quiet this morning. Thank you, alright.

A

So, as as Liz mentioned, there's been some slight changes at core OS I have a new email address, so good and I I didn't want to focus on this, but just wanted to point it out that core OS was acquired by Red Hat and really the core OS team continues to remain focus on the kubernetes ecosystem of a lot of the open source projects that we introduced and if you want to learn more about it, there's Red Hat summit next week and I'm gonna spend my time at coupon talking about open source and convene eddie's.

A

So two years ago, corliss introduced this concept of operators and luckily Aparna gave a live. Demo. I only have five minutes, so I can do a live demo, but without coordination we got a live demo out of it anyways and so operators. This idea of kind of these cube negative applications that are managed via the kubernetes api and then also deployed on the api, and so they run in pods they're managed with cube api's here's the demo. I don't have to do this.

A

I can save so much time all right, so so the analogy here is imagine a cloud provider has all these different services have security services, identity services? It runs on servers if you didn't know that about the cloud that servers and what we're trying to do here with operators is imagine a cloud where you can install new applications on it. Most of the public cloud providers are not going to. Let you just your application. Your your WordPress server make a new API to create new instances of your of your WordPress application or your enterprise application.

A

That's that's simply not possible on the on the public cloud, but with kubernetes we can extend kubernetes and add whatever application, we'd like whether it be spark or at City or prometheus or maybe applications that you find important inside of your organization and so own, introduce it two years ago and where we are today is a lot of people have found a lot of value out of these things.

A

This is from Mike goodness a Ticketmaster which they introduced Prometheus operator into their clusters at Ticketmaster, allowing application, engineers and sres to manage their own monitoring systems for their applications, not infrastructure of monitoring but application monitoring and it kind of took off fast, hundreds of Prometheus instances and their clusters hooked up to various alerting, etc. So, clearly, this pattern is useful.

A

It's always good to see when you introduce a simple concept and then it explodes and use that know, that's how you know: something's good and so building on that we wanted to ask ourselves: how do we create more operators easily and then how do we bring operators to more users?

A

So, a couple days ago we introduced the operator framework. It was an open source project and I'm happy to announce it here at coop con, and what we're trying to do is twofold actually tackle those two problems. It's here at github, comm, slash operator framework. You can check it out, there's a bunch of stuff in there we're introducing a few more components over the next few weeks, but the things that we've open-sourced so far.

A

First is the operator SE k, which is tools to build package and test your operator and there's a bunch of challenges: building these operators they're non-trivial pieces of software tracking, related cube resources, tests, scaffolding, how do I vendor the correct, kubernetes, libraries etc, and the operator sdk has a command-line tool that generates the scaffold for you and then a bunch of simple what simplified ap is to interact with the community system and we have kind of shared this mission in parallel, we've been building these things, Philip Roth Google engineer has been building something called COO builder and he has this great analogy.

A

What we're trying to do is Dooku be on Rails if you Ruby on Rails and so we're trying to simplify the application development on top of kubernetes to build more of these operators, because people are getting so much value out of them. The other piece is lifecycle management, so in a partner's demo, you saw that it was download some stuff from github like a gamma file, and then you deploy the CMO file, which is useful for demos, but not really useful in production, and so we want to answer these questions.

A

How do we maintain a catalogue? How do we make of operators how they make it easy to upgrade? How do we make it easy to insert these into different namespaces, and so we kind of take these manifests this catalogue and install them into namespaces and it's kind of handling this arrow? That is inserting the application into your cloud. Alright, you can handle built-in types, helm, charts or things built with the operator SDK and that's the recap: we're open sourcing.

A

This thing you can find it up on github we're trying to build more operators more easily for more users and the. Why is we want cuban IDs to be the dominant api for these cloud? Apps we're around cube con Jimmy, Robb, myself and Kelly. If you want to find us we'll be at the Red Hat booth thanks a lot.