MicroK8s is Lightweight and pure upstream K8s. The smallest, simplest, pure production K8s. For clusters, laptops, IoT and Edge, on Intel and ARM.

This installs MicroK8s using Github Actions.

More information about MicroK8s

This Github Action only works on Linux machines.

This Github Actions enables one to test their applications on multiple Kubernetes versions by following MicroK8s channels.

Required This is the MicroK8s channel to choose. Example: latest/stable or 1.26/stable

Strict confinement is also available see

New from v0.2.0

Optional a JSON array containing the MicroK8s addon to enable. Example addons: "['prometheus','metrics-server','linkerd']"

Removed from v0.3.0

Deprecated in favor of the new argument addons

Optional Since MicroK8s does not enable RBAC by default, user can choose whether they want to enable rbac or not.

Removed from v0.3.0

Deprecated in favor of the new argument addons

Optional Since MicroK8s does not enable dns by default, user can choose whether they want to enable CoreDNS or not.

Removed from v0.3.0

Deprecated in favor of the new argument addons

Optional Since MicroK8s does not enable storage by default, user can choose whether they want to enable local hostPath storage or not.

New from v0.3.0

New from v0.4.0

Optional A launch configuration is a YAML configuration file that can be applied to a MicroK8s node typically during installation. Example: /home/runner/microk8s-config.yaml

Once you have defined your launch-configuration, there is no need to add the addons one by one in the config.

Example:

- uses: balchua/microk8s-actions@feat/launch_config
  with:
    channel: "1.28/stable"
    launch-configuration: "$GITHUB_WORKSPACE/microk8s-config.yaml"
    sideload-images-path: "$HOME/sideload/"

This is only available from MicroK8s v1.27

New from v0.4.0

Optional

Side load images automatically loads image tarballs to MicroK8s Containerd.
You can use this to automatically load your images into MicroK8s, without using a registry .

Example: /home/runner/sideload/

Please note that this is applicable during MicroK8s installation.

This is only available from MicroK8s v1.27

Below shows how one can use the Action.

name: Test Microk8s
on: [push]

jobs:
  test:
    runs-on: ubuntu-latest
    name: A job to install MicroK8s
    steps:
    - uses: balchua/[email protected]
      with:
        channel: '1.26/stable'
        addons: '["dns", "rbac", "hostpath-storage", "registry"]'
    - name: Test MicroK8s
      id: myactions
      run: |
        kubectl get no
        kubectl get pods -A -o wide
        

To use microk8s commands for example: microk8s status, you should use sudo or sg command.

    - name: Test MicroK8s commands
      id: microk8s
      run: |
        set -ex
        echo "Executing microk8s status using sg command."
        sg microk8s -c 'microk8s status'
        echo "Executing microk8s status using sudo command."
        sudo microk8s status

sg commands will execute the command using the group specified. Description taken from here

The sg command works similar to newgrp but accepts a command. The command will be executed with the /bin/sh shell. With most shells you may run sg from, you need to enclose multi-word commands in quotes. Another difference between newgrp and sg is that some shells treat newgrp specially, replacing themselves with a new instance of a shell that newgrp creates. This doesn't happen with sg, so upon exit from a sg command you are returned to your previous group ID.

From v0.3.0 this action now supports Snap strict confinement.

Please note that MicroK8s strict confinement is only available from 1.25 as a separate track 1.25-strict.

More information here

jobs:
  test:
    runs-on: ubuntu-latest
    name: A job to install MicroK8s with strict confinement
    steps:
    - uses: balchua/[email protected]
      with:
        channel: '1.26-strict/stable'
        devMode: 'true'
        addons: '["rbac", "dns", "hostpath-storage", "registry", "metrics-server"]'

Removed from v0.3.0, since it is no longer part of MicroK8s addon.

You can use the kubeflow microk8s addon to test your ML. Due to the size of the VM used by github actions the KUBEFLOW_BUNDLE used is edge. This is the appropriate bundle for low resource machines.

For more information regarding kubeflow addon check this link

The main program is a Typescript, located in src.

Before pushing the code to Github, you should compile the source to ES6.

$ npm run build

See Troubleshooting guide