Red Hat OpenShift Spain 2022 OpenShift Commons Gathering at KubeconEU, 17 May 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Declarative GitHub org Management with Peribolos on Operate First - Tom Coufal (Red Hat) OSCG 2022

Description

Declarative GitHub org Management with Peribolos on Operate First
Speaker: Tomáš Coufal (Red Hat)

OpenShift Commons Gathering Kubecon EU
May 17, 2022 Live from Kubecon EU in Valencia, Spain

Full Agenda here: https://commons.openshift.org/gatherings/OpenShift_Commons_Gathering_at_Kubecon_Europe_2022.html

Learn more at: https://commons.openshift.org

A

Hello, everybody welcome to my lightning talk about experiments. uh We run with services and a very quick, very short story of uh what we can do with services beyond contributing to to open source uh projects. So I want to share share a story about uh creating services in a world of software as a service, and uh what does it mean for open source tooling?

A

Today, I'm going to be talking about peripherals on operate first, which is something we are trying to shape and build as something called open service. So what exactly? Does that mean to have an open service um when you think of a current ecosystem in open source um people contribute to tools to to software code bases uh and all these bits and pieces that make up a successful application?

A

But when you want to contribute beyond that and see how applications are actually run and executed, and the life cycle is taken care of there's not much ways to contribute to that, so we're trying to change that through uh experiments like this one.

A

This is a story of privilege exactly three years ago on kubecon 2019 there's this talk called preblos how kubernetes uses githubs to manage github communities at scale.

A

This is a then it was a brand new tool called privilege. Now it's part of the kubernetes community, tooling, through pro and and through through other ci means, and it allows you to manage github organizations as a code.

A

We thought this is a very neat idea. So how does that work? You get a config file that you can do changes to uh it's a declarative way to to describe your github organization right. So you can add members, you can add, create repositories, you can create teams uh and what's more important, uh people can do that. You are basically crowdsourcing the github organization management. So you don't need to hunt down your github organization admin just to add a member to it.

A

They can create a pull request. You can review it and once merged it gets applied to github. So we thought that this is a neat idea and that we want uh to use that in our organization.

A

So how would you do that? Traditionally? Normally, you would start with taking the tool uh taking either a binary version of it or a container activation version of it, pull it into your ci and let the ci run on your repository, where you have version controlled this configuration file. This is a normal workflow that probably most of us would use well that at some point becomes a problem when you're managing multiple organizations.

A

You need to set this thing up for every organization. You need to maintain and upgrade these peripherals applications across all these organizations made in the same configuration and so on and so on, and then it becomes a challenge and then it becomes a challenge that you need to take care of life cycle of all those ci pipelines and manage it all. Similarly, hopefully, the same, but every organization is a bit different and usually these code bases diverge. So how how to solve that?

A

um Well, nowadays, a trendy thing to do: let's make it a service right, that's what everybody does with everything currently in the sas ecosystem: let's create a service that is managing all of these organizations. Together, we thought okay, let's do that.

A

So we created this github application that can that you can sign up for install into your organization and let it manage your organization through your configuration file inside your organization, which is a normal thing, that modern application would use, create a service out of your tools out of things that that you use in this decentralized fashion and unified into a service. But we thought can we make more out of it?

A

A

How can we get more value out of a service? So we decided to uh make it make it open, uh make it a collaborative way? Let's challenge, we decided to challenge ourselves and uh collaboratively manage this service as a community, so we found uh a community where we can. We can do so for various services, not just for our little tiny, uh marginal application that takes care of of a single use case.

A

So we joined this operate first community, where, which is a community supported by red hat and its focus is exactly this find a common ground find a place, a meeting spot where we can talk about how ops should be done, how sre work should be done? How to handle services collaboratively and how to transparently contribute to something which is running as a managed offering.

A

So we join this community and we use it to share best practices, define best practices for managed services and to also host our our workload because operate. First community has a community cloud offering that is publicly available for anybody to join and use, and it's basically free hardware that you can that you can jump in and use and and deploy your workloads onto.

A

So we decided that this is the best place where we can run our our service.

A

And by that, I'm going to conclude this this short story: you can join us in operate. First, you can join us in the peribos application. um We would welcome any contributions to either of those things uh collaborative code management credibility's, an application uh install it use it. Thank you very much.

A