This GitHub Action is based on koslib/helm-eks-action with the improvement that it runs on aws-cli-v2 as opposed to aws-cli-v1.
The functionality and the usage is same.
I did not have enough time to make a PR and get it merged so I forked the original action.
The action is almost fully compatible with the original plugin.
Github Action for executing Helm commands on EKS .
The Helm version installed is Helm3.
This action was inspired by kubernetes-action.
This Github Action was created with EKS in mind, therefore the following example refers to it.
plugins
: you can specify a list of Helm plugins you'd like to install and use later on in your command. eg. helm-secrets or helm-diff. This action does not support only a specific list of Helm plugins, rather any Helm plugin as long as you supply its URL. You can use the following example as a reference.command
: your kubectl/helm command. This supports multiline as per the Github Actions workflow syntax.
example for multiline:
...
with:
command: |
helm upgrade --install my-release chart/repo
kubectl get pods
name: deploy
on:
push:
branches:
- master
- develop
jobs:
deploy:
runs-on: ubuntu-latest
env:
AWS_REGION: us-east-1
CLUSTER_NAME: my-staging
steps:
- uses: actions/checkout@v3
- name: AWS Credentials
uses: aws-actions/configure-aws-credentials@v1
with:
role-to-assume: arn:aws:iam::<your account id>:role/github-actions
role-session-name: ci-run-${{ github.run_id }}
aws-region: ${{ env.AWS_REGION }}
- name: kubeconfig
run: |
aws eks update-kubeconfig --name ${{ env.CLUSTER_NAME }} --region ${{ env.AWS_REGION }} --kubeconfig ./kubeconfig
echo 'KUBE_CONFIG_DATA<<EOF' >> $GITHUB_ENV
echo $(cat ./kubeconfig | base64) >> $GITHUB_ENV
echo 'EOF' >> $GITHUB_ENV
- name: helm deploy
uses: harsh098/helm-eks-action-awscli-2@v0.3
env:
KUBE_CONFIG_DATA: ${{ env.KUBE_CONFIG_DATA }}
with:
plugins: "https://github.com/jkroepke/helm-secrets" # optional
command: helm secrets upgrade <release name> --install --wait <chart> -f <path to values.yaml>
Use the output of your command in later steps
steps:
- name: Get URL
id: url
uses: harsh098/helm-eks-action-awscli-2@v0.3
with:
command: kubectl get svc my_svc -o json | jq -r '.status.loadBalancer.ingress[0].hostname'
- name: Print Response
run: echo "Response was ${{ steps.url.outputs.response }}"
The latest version of this action uses the following dependencies versions:
Package | Version |
---|---|
awscli | v2.11.25 |
helm | v3.3.0 |
It is very much possible that an update came out and I did not update the action on time. In this please, feel free to send me a PR and I'll review it as soon as possible.
It is required to set the KUBE_CONFIG_DATA
env/secret in order to access your cluster. I recommend you do it dynamically using a step like that:
- name: kubeconfig
run: |
aws eks update-kubeconfig --name ${{ env.CLUSTER_NAME }} --region ${{ env.AWS_REGION }} --kubeconfig ./kubeconfig
echo 'KUBE_CONFIG_DATA<<EOF' >> $GITHUB_ENV
echo $(cat ./kubeconfig | base64) >> $GITHUB_ENV
echo 'EOF' >> $GITHUB_ENV
If you find this configuration option complicated, you can still supply KUBE_CONFIG_DATA
as a repository secret, however this is not endorsed by this repository.
Pull requests, issues or feedback of any kind are more than welcome by anyone!
If this action has helped you in any way and enjoyed it, feel free to submit feedback through issues