Cloud Native Computing Foundation KubeCon + CloudNativeCon Europe 2019 (Barcelona), 22 May 2019

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Sponsored Keynote: Expanding the Kubernetes Operator Community - Rob Szumski

Description

Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io

Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects

Sponsored Keynote: Expanding the Kubernetes Operator Community - Rob Szumski, Principal Product Manager for OpenShift, Red Hat

Operators enable you run those “tricky” stateful services and complex distributed systems on Kubernetes. Kubernetes Operators are embedded with all of the special knowledge for running an app, from the experts in the open source community, so that every end user doesn’t need to be one. Discover Operators ready to use on your cluster with OperatorHub.io and learn about the tools at your disposal to build, test and ship an Operator.

https://sched.co/MRyx

A

Everybody today, I want to give you an update on the expanding operator ecosystem.

A

The operator community kind of birthed out of this idea that you know running stateless applications is pretty easy to do on Cube and we got the staple applications and then really quickly. We got to these advanced distributed systems running on kubernetes, and this is really great, but there aren't primitives built in to kubernetes to handle these types of applications.

A

If you want to do data, rebalancing or seamless upgrades, and so that's where an operator enters, the picture and operators run these complex applications on your behalf and the way that they work is you take the experts in an application how its installed, how its failed over how its upgraded and they take that expertise and embed it into the operator such that you take those best practices and your customizations of the application, which might be different in staging or production and then translate those into kubernetes objects that run on the cluster.

A

This is really really powerful, because you're using kubernetes primitives in the toolkit that you already get thanks to all the hard work upstream, so you're not reinventing that and you're, focusing on your application and because you're focusing on kubernetes you get something that is truly hybrid. You can take this operator and run the application on any kubernetes cluster out there move it around, as you would like. The operator framework is a set of tools that can help you build these operators, test them and deploy them in production.

A

It's really useful to look at an example of these to really drive at home. So let's just say that we want to run Postgres and we've got a number of options for this. We can run Postgres in a container on our laptop. That's really easy.

A

We could also go pull off a cloud database and get a bunch of nice features with that or we could use the crunchy data operator that runs Postgres for us and, as you can see, the crunchy data operator gives you that same cloud like experience, but we have a few special extra superpowers. One being that we can run this locally, I can run a full stack on my laptop on a plane to do some development or it's open source, so you can fully introspect what's going on. This is especially helpful when you want to go.

A

Look at the interactions between some of the components: go debug some of the pods if you're running into an issue- and this is really really really powerful, and so what this gives you is this extreme transparency into. What's go on so here we're looking at some screenshots of a Kafka cluster, and you can see some of the environment variables that were setting here. You can see different versions of the components and then we can get streaming logs, because this is just running on our cluster.

A

We've got full control over what we can see, and this is really nice for you to see exactly what's going on under the hood, so I went and did some digging and these tweets came across my timeline, and these are some folks that are actually finding bugs in public cloud services and what's interesting about this, the gentleman on the Left spent 3 entire weeks debugging an application I thought it was his bug and actually turns out it's a bug in this cloud database and so he's wasted three weeks.

A

It's really great that the bug was fixed in this case, but you know we could save him a bunch of time by actually exposing a little bit more information about. What's actually going on under the hood and on the right hand, side. This gentleman has actually thinks he found a bug in a cloud database and he's just gonna write some software to work around that bug, but in a better model couldn't he actually verify that this is the case and then let's actually try to get this bug fixed instead of writing a software workaround.

A

So this is kind of what you want to go for you want something: that's a little bit more transparent. You can actually see what's going on and it saves everybody a whole bunch of time.

A

So the crunchy data operator we just looked at is one of a number on operator hub do, and this is a community listing of operators, we're adding new operators every single day and so some of the new ones that just came in the elasticsearch operator from elastic a bunch of the come framework components of Sto the service mesh or in there.

A

So take a look, there's a whole bunch more out there really really great resource, and then, if this, this vision of a true hybrid cloud, this cloud, like experience coming to your kubernetes cluster, resonates with you or your customers. Red Hat wants to help you bring these products to market. We have a new operator certification program and this is really important because Red Hat is truly neutral in this.

A

You know there's some open source licensing shenanigans going on right now, and you know you need to pick a partner, that's truly neutral and wants to bring this to market with you, so that interest you please get in touch. We would love to have you now once you're an admin and you've got these operators installed on your cluster. You get full control over. What's going on so here you can introspect the versions of the operators that we have running see what's namespaces those capabilities are exposed into. So you really have really nice visibility of.

A

What's going on, so you can run your production infrastructure, but the most important thing is: this is all self service for your engineers. These are the folks that are using your cluster they're running the applications, and so they have GUI access. You can deploy these via the command line, and you know you're just eating into your cluster quota. You don't have to call anybody to change an environment variable and staging and promote that to production of great things, and it's really really powerful for speeding up your team's.

A

So I want to invite you to check out operator hub, do and take a look at some of these great operators. But, more importantly, if you want to build an operator yourself, take a look at our getting started guide for our sdk. It's really really quick and easy to get started and I want to encourage you not just to focus on the partners that you saw on stage today, but building operators for your internal applications. Everybody's got really complex state going on. You need to share that between groups even inside of your company.

A

So an SDK is a really great way to get started with an operator to do that. And then, if you choose to do this- and you want to you, know, run some common scenarios by other folks solve some problems. Collaboratively we've got an operator sig that meets monthly and we love to have you join that community as well. So that's a quick update on the operator ecosystem. Thank you. So much.