A Kubernetes Operator to put Darkroom on Autopilot
Darkroom is a great image proxy to serve your images from your desired source and perform image manipulations on the fly.
This operator aims to make to it easy to deploy Darkroom in a Kubernetes Cluster and make it easy to manage the cluster with an intuitive GUI.
TBD
Install Operator SDK v1.5.0
brew install operator-sdkNote: We recommend going through operator-sdk getting started guide if you are new to operator development
Operator SDK uses etcd bundled with KubeBuilder, install it with bin/install-kubebuilder.
[Optional] Add
/usr/local/kubebuilder/binto your $PATH
The structure of this project is different than other operator codebases you might be familiar with. Since this project also includes an API Server with GUI and makes use of same API definitions, we have kept it this way.
To generate a new API object, run the command
operator-sdk create api --group deployments --version v1alpha1 --kind Darkroom --resource=true --controller=trueThis will generate the required resource definition and controller under ./api/ and ./controllers/, move these to ./pkg/api/ and ./internal/controllers/ respectively. But the command will fail as there is no root level main.go file.
The changes required in main.go are hence skipped and you must add the generated resource scheme to runtime scheme in cmd/operator/cmd/root.go and setup the Reconciler with the Controller Manager manually.
You can run make operator/generate to generate necessary code by the controller-gen. And run make operator/manifests to generate the required YAML definitions.
It is recommended to use a Kind cluster for faster iteration.
Install cert-manager with
kubectl apply -f https://github.com/jetstack/cert-manager/releases/download/v1.2.0/cert-manager.yamlNote: You may use the latest version
Read our contributing guide to learn about our development process, how to propose bugfixes and improvements, and how to build and test your changes to Darkroom Operator.
Darkroom Operator is MIT licensed.