Contributors and newcomers to the Jenkins infrastructure project do not want to be confused: coherency is expected on the conventions chosen for our repositories.
The repository https://github.com/jenkins-infra/charts defines 2 different Kubernetes clusters (as for today):
- `publick8s`, an AKS cluster used for public and privates web services + a few private Jenkins controllers, defined here: https://github.com/jenkins-infra/charts/blob/master/clusters/publick8s.yaml
- `cik8s`, an EKS cluster only used to provide build capacity in the form of Kubernetes Pod agents for the Jenkins controller at https://ci.jenkins.io, defined here: https://github.com/jenkins-infra/charts/blob/master/clusters/cik8s.yaml
Each of this cluster have 2 types of applications: some are common to all our cluster (existing and upcoming), others are specific to a given cluster. It means that you can find 3 directories to host the application's configurations at https://github.com/jenkins-infra/charts/tree/master/config: `cik8s`, `publick8s` and `default`.
However, almost all the configuration of `publick8s` are defined inside the `default` directory: https://github.com/jenkins-infra/charts/tree/master/config/default , which is not coherent.
- Open a PR that would move the correct config YAML files into the correct directories
- A PR had been opened, review, approved and merged
- Each config file common for both clusters must resides in the `config/default`, while the others should be on their respective specific cluster
- At least one successfull deployment done on the Kubernetes clusters (includes fixing any breaking application)