e2e.yaml

name: e2e

on:
  workflow_dispatch:
  push:
    branches-ignore:
    - master
    tags-ignore: [ '*' ]

jobs:
  run-on-kubernetes:
    runs-on: ubuntu-latest
    steps:
      - name: Checkout
        uses: actions/checkout@692973e3d937129bcbf40652eb9f2f61becf3332 # v4
      # Seemingly something mise installs breaks post-checkout.
      # this runs a post-task after post-mise (for caching) but before post-checkout
      - name: Clean up Mise
        uses: gacts/run-and-post-run@v1
        with:
          run: echo "Nothing"
          post: mise implode -y
      - name: Install tools
        uses: jdx/mise-action@v2
      - name: Create a test cluster and deploy this branch
        run: ./tests/run_cluster_in_kind.sh ${GITHUB_REF#refs/heads/}
      - name: Test resources come up
        run: ./tests/run_tests.sh
      - name: Debug failure summaries
        if: failure()
        run: |
          echo "echo "All Flux components""
          flux get all --all-namespaces
          echo "Helm list"
          helm list -A
          echo "kubectl get all in namespace flux-system"
          kubectl -n flux-system get all -o wide
          echo "kubectl get all helm releases"
          kubectl get helmreleases.helm.toolkit.fluxcd.io --all-namespaces
          echo "kubectl get all pods"
          kubectl get pods -A -o wide
      - name: Debug describe nodes
        if: failure()
        run: |
          kubectl describe nodes
      - name: Debug describe pods
        if: failure()
        run: |
          kubectl describe pods --all-namespaces
      - name: Debug describe helm releases
        if: failure()
        run: |
          kubectl describe helmreleases.helm.toolkit.fluxcd.io --all-namespaces
      - name: Debug source-controller logs
        if: failure()
        run: |
          kubectl -n flux-system logs deploy/source-controller
      - name: Debug kustomize-controller logs
        if: failure()
        run: |
          kubectl -n flux-system logs deploy/kustomize-controller
      - name: Debug helm-controller logs
        if: failure()
        run: |
          kubectl -n flux-system logs deploy/helm-controller
      - name: Debug kwok controller
        if: failure()
        run: |
          kubectl -n kube-system logs deploy/kwok-controller
          kubectl -n kube-system logs deploy/kwok-controller --previous