efrecon/vcluster repository overview

⁠vcluster in Docker

This Docker image⁠ provides the vcluster⁠ CLI in a Docker container. This image also has helm⁠ installed, as it is a required dependency. It is automatically generated⁠ from of a project⁠ for building kubernetes-oriented Docker images.

The image provides a number of build-time arguments that can be changed:

• HELM_VERSION should be an existing version of the helm client (with or without the leading v).
• VCLUSTER_VERSION can either be the string latest, or an existing version (with or without the leading v). When latest, the latest stable version at build-time will be installed, otherwise the version matching the value of the build-time argument, if it exists.

vcluster is the entrypoint of the image. To use the image, you will probably have to mount your kubeconfig file and the default helm directories (alt. declare these through HELM_CACHE_HOME, HELM_CONFIG_HOME and HELM_DATA_HOME), e.g.

docker run -it --rm \
  -u $(id -u):$(id -g) \
  -v ${HOME}/.kube/config:/.kube/config:ro \
  -v ${HOME}/.cache/helm:/.cache/helm \
  -v ${HOME}/.config/helm:/.config/helm \
  -v ${HOME}/.local/share/helm:/.local/share/helm \
  efrecon/vcluster help

Everytime a new stable version of vcluster is made available, this image will automatically be rebuilt. At the time of the build, the latest version of helm will be detected and the build will generate an image that has both binaries. The same kind of dependency exists towards kubectl, which is also installed in the image. This means that every time the image is rebuilt, the latest of helm will be used. This is a known side-effect of the build process and might not always be what you want.