conformance-aks.yaml

name: Conformance AKS (ci-aks)

# Any change in triggers needs to be reflected in the concurrency group.
on:
  workflow_dispatch:
    inputs:
      PR-number:
        description: "Pull request number."
        required: true
      context-ref:
        description: "Context in which the workflow runs. If PR is from a fork, will be the PR target branch (general case). If PR is NOT from a fork, will be the PR branch itself (this allows committers to test changes to workflows directly from PRs)."
        required: true
      SHA:
        description: "SHA under test (head of the PR branch)."
        required: true
      extra-args:
        description: "[JSON object] Arbitrary arguments passed from the trigger comment via regex capture group. Parse with 'fromJson(inputs.extra-args).argName' in workflow."
        required: false
        default: '{}'
  push:
    branches:
      - 'renovate/main-**'
  # Run every 6 hours
  schedule:
    - cron:  '0 0/6 * * *'

# By specifying the access of one of the scopes, all of those that are not
# specified are set to 'none'.
permissions:
  # To read actions state with catchpoint/workflow-telemetry-action
  actions: read
  # To be able to access the repository with actions/checkout
  contents: read
  # To allow retrieving information from the PR API
  pull-requests: read
  # To be able to set commit status
  statuses: write
  # Required to generate OIDC tokens for `az` authentication
  id-token: write

concurrency:
  # Structure:
  # - Workflow name
  # - Event type
  # - A unique identifier depending on event type:
  #   - schedule: SHA
  #   - workflow_dispatch: PR number
  #
  # This structure ensures a unique concurrency group name is generated for each
  # type of testing, such that re-runs will cancel the previous run.
  group: |
    ${{ github.workflow }}
    ${{ github.event_name }}
    ${{
      (github.event_name == 'push' && github.sha) ||
      (github.event_name == 'schedule' && github.sha) ||
      (github.event_name == 'workflow_dispatch' && github.event.inputs.PR-number)
    }}
  cancel-in-progress: true

env:
  name: ${{ github.repository_owner }}-${{ github.event.repository.name }}-${{ github.run_id }}-${{ github.run_attempt }}
  cost_reduction: --node-vm-size Standard_B2s --node-osdisk-size 30
  cilium_cli_ci_version:

jobs:
  echo-inputs:
    if: ${{ github.event_name == 'workflow_dispatch' }}
    name: Echo Workflow Dispatch Inputs
    runs-on: ubuntu-22.04
    steps:
      - name: Echo Workflow Dispatch Inputs
        run: |
          echo '${{ tojson(inputs) }}'

  commit-status-start:
    name: Commit Status Start
    runs-on: ubuntu-latest
    steps:
      - name: Set initial commit status
        uses: myrotvorets/set-commit-status-action@3730c0a348a2ace3c110851bed53331bc6406e9f # v2.0.1
        with:
          sha: ${{ inputs.SHA || github.sha }}

  generate-matrix:
    name: Generate Matrix
    runs-on: ubuntu-latest
    outputs:
      matrix: ${{ steps.set-matrix.outputs.matrix }}
      empty: ${{ steps.set-matrix.outputs.empty }}
    steps:
      - name: Checkout context ref (trusted)
        uses: actions/checkout@692973e3d937129bcbf40652eb9f2f61becf3332 # v4.1.7
        with:
          ref: ${{ inputs.context-ref || github.sha }}
          persist-credentials: false

      - name: Convert YAML to JSON
        run: |
          work_dir=".github/actions/azure"
          destination_directory="/tmp/generated/azure"
          mkdir -p "${destination_directory}"

          yq -o=json "${work_dir}/k8s-versions.yaml" | jq . > "${destination_directory}/azure.json"

      - name: Generate Matrix
        run: |
          cd /tmp/generated/azure

          # Use complete matrix in case of scheduled run
          # main -> event_name = schedule
          # other stable branches -> PR-number starting with v (e.g. v1.14)
          if [[ "${{ github.event_name }}" == "schedule" || "${{ inputs.PR-number }}" == v* ]];then
            jq '{ "include": [ .include[] | select(.disabled==null) ] }' azure.json > /tmp/matrix.json
          else
            jq '{ "include": [ .include[] | select(.default) ] }' azure.json > /tmp/matrix.json
          fi

          echo "Generated matrix:"
          cat /tmp/matrix.json

      - name: Login to Azure
        uses: azure/login@6c251865b4e6290e7b78be643ea2d005bc51f69a # v2.1.1
        with:
          creds: ${{ secrets.AZURE_PR_SP_CREDS }}

      - name: Filter Matrix
        id: set-matrix
        run: |
          cp /tmp/matrix.json /tmp/result.json
          jq -c '.include[]' /tmp/matrix.json | while read i; do
            VERSION=$(echo $i | jq -r '.version')
            LOCATION=$(echo $i | jq -r '.location')
            az aks get-versions --location $LOCATION > /tmp/output
            if grep -q -F $VERSION /tmp/output; then
              echo "Version $VERSION is valid for location $LOCATION"
            else
              echo "::notice::Removing version $VERSION as it's not valid for location $LOCATION"
              jq 'del(.include[] | select(.version == "'$VERSION'"))' /tmp/result.json > /tmp/result.json.tmp
              mv /tmp/result.json.tmp /tmp/result.json
            fi
          done
          echo "Filtered matrix:"
          cat /tmp/result.json
          echo "matrix=$(jq -c . < /tmp/result.json)" >> $GITHUB_OUTPUT
          echo "empty=$(jq '(.include | length) == 0' /tmp/result.json)" >> $GITHUB_OUTPUT

  installation-and-connectivity:
    name: Installation and Connectivity Test
    needs: generate-matrix
    if: ${{ needs.generate-matrix.outputs.empty == 'false' }}
    runs-on: ubuntu-latest
    timeout-minutes: 90
    env:
      job_name: "Installation and Connectivity Test"
    strategy:
      fail-fast: false
      matrix: ${{fromJson(needs.generate-matrix.outputs.matrix)}}

    steps:
      - name: Collect Workflow Telemetry
        uses: catchpoint/workflow-telemetry-action@94c3c3d9567a0205de6da68a76c428ce4e769af1 # v2.0.0
        with:
          comment_on_pr: false

      - name: Checkout context ref (trusted)
        uses: actions/checkout@692973e3d937129bcbf40652eb9f2f61becf3332 # v4.1.7
        with:
          ref: ${{ inputs.context-ref || github.sha }}
          persist-credentials: false

      - name: Set Environment Variables
        uses: ./.github/actions/set-env-variables

      - name: Override cluster name
        run: |
          # Extend default name with matrix index to avoid cluster name conflicts
          NAME=${{ env.name }}-${{ matrix.index }}
          echo "name=${NAME}" >> "$GITHUB_ENV"

      - name: Get Cilium's default values
        id: default_vars
        uses: ./.github/actions/helm-default
        with:
          image-tag: ${{ inputs.SHA }}
          chart-dir: ./untrusted/install/kubernetes/cilium

      - name: Set up job variables
        id: vars
        run: |
          if [ "${{ github.event_name }}" = "workflow_dispatch" ]; then
            OWNER="${{ inputs.PR-number }}"
          else
            OWNER="${{ github.ref_name }}"
            OWNER="${OWNER//[.\/]/-}"
          fi

          CILIUM_INSTALL_DEFAULTS="${{ steps.default_vars.outputs.cilium_install_defaults }} \
            --helm-set cluster.name=${{ env.name }} \
            --helm-set loadBalancer.l7.backend=envoy \
            --helm-set tls.secretsBackend=k8s \
            --helm-set=azure.resourceGroup=${{ env.name }} \
            --helm-set ipam.operator.clusterPoolIPv4PodCIDRList=192.168.0.0/16" # To avoid clashing with the default Service CIDR of AKS (10.0.0.0/16)
          CONNECTIVITY_TEST_DEFAULTS="--flow-validation=disabled \
            --hubble=false --collect-sysdump-on-failure --external-target bing.com. --external-cidr 8.0.0.0/8 --external-ip 8.8.4.4 --external-other-ip 8.8.8.8"
          echo cilium_install_defaults=${CILIUM_INSTALL_DEFAULTS} >> $GITHUB_OUTPUT
          echo connectivity_test_defaults=${CONNECTIVITY_TEST_DEFAULTS} >> $GITHUB_OUTPUT
          echo sha=${{ steps.default_vars.outputs.sha }} >> $GITHUB_OUTPUT
          echo owner=${OWNER} >> $GITHUB_OUTPUT

      - name: Login to Azure
        uses: azure/login@6c251865b4e6290e7b78be643ea2d005bc51f69a # v2.1.1
        with:
          creds: ${{ secrets.AZURE_PR_SP_CREDS }}

      - name: Create AKS cluster
        run: |
          # Create group
          az group create \
            --name ${{ env.name }} \
            --location ${{ matrix.location }} \
            --tags usage=${{ github.repository_owner }}-${{ github.event.repository.name }} owner=${{ steps.vars.outputs.owner }}

          # Create AKS cluster
          az aks create \
            --resource-group ${{ env.name }} \
            --name ${{ env.name }} \
            --location ${{ matrix.location }} \
            --kubernetes-version ${{ matrix.version }} \
            --network-plugin none \
            --node-count 2 \
            ${{ env.cost_reduction }} \
            --generate-ssh-keys

      - name: Get cluster credentials
        run: |
          az aks get-credentials \
            --resource-group ${{ env.name }} \
            --name ${{ env.name }}

      - name: Install Cilium CLI
        uses: cilium/cilium-cli@e386af2b9f500e4c40436ac660cd6602da104fc7 # v0.16.14
        with:
          skip-build: ${{ env.CILIUM_CLI_SKIP_BUILD }}
          image-repo: ${{ env.CILIUM_CLI_IMAGE_REPO }}
          image-tag: ${{ env.CILIUM_CLI_VERSION }}

      - name: Wait for images to be available
        timeout-minutes: 30
        shell: bash
        run: |
          for image in cilium-ci operator-azure-ci hubble-relay-ci ; do
            until docker manifest inspect quay.io/${{ env.QUAY_ORGANIZATION_DEV }}/$image:${{ steps.vars.outputs.sha }} &> /dev/null; do sleep 45s; done
          done

      # Warning: since this is a privileged workflow, subsequent workflow job
      # steps must take care not to execute untrusted code.
      - name: Checkout pull request branch (NOT TRUSTED)
        uses: actions/checkout@692973e3d937129bcbf40652eb9f2f61becf3332 # v4.1.7
        with:
          ref: ${{ steps.vars.outputs.sha }}
          persist-credentials: false
          path: untrusted
          sparse-checkout: |
            install/kubernetes/cilium

      - name: Install Cilium
        id: install-cilium
        run: |
          cilium install ${{ steps.vars.outputs.cilium_install_defaults }}

      - name: Enable Relay
        run: |
          cilium hubble enable

      - name: Wait for Cilium status to be ready
        run: |
          cilium status --wait --wait-duration=10m

      - name: Make JUnit report directory
        run: |
          mkdir -p cilium-junits

      - name: Run connectivity test (${{ join(matrix.*, ', ') }})
        run: |
          cilium connectivity test ${{ steps.vars.outputs.connectivity_test_defaults }} \
          --junit-file "cilium-junits/${{ env.job_name }} (${{ join(matrix.*, ', ') }}) - 1.xml" \
          --junit-property github_job_step="Run connectivity test (${{ join(matrix.*, ', ') }})"

      - name: Clean up Cilium
        run: |
          cilium uninstall --wait

      - name: Create custom IPsec secret
        run: |
          kubectl create -n kube-system secret generic cilium-ipsec-keys --from-literal=keys="15 rfc4106(gcm(aes)) $(echo $(dd if=/dev/urandom count=20 bs=1 2> /dev/null | xxd -p -c 64)) 128"

      - name: Install Cilium with encryption
        run: |
          cilium install ${{ steps.vars.outputs.cilium_install_defaults }} \
            --helm-set encryption.enabled=true \
            --helm-set encryption.type=ipsec

      - name: Enable Relay
        run: |
          cilium hubble enable

      - name: Wait for Cilium status to be ready
        run: |
          cilium status --wait --wait-duration=10m

      - name: Run connectivity test with IPSec (${{ join(matrix.*, ', ') }})
        run: |
          cilium connectivity test ${{ steps.vars.outputs.connectivity_test_defaults }} --force-deploy \
          --junit-file "cilium-junits/${{ env.job_name }} (${{ join(matrix.*, ', ') }}) - 2.xml" \
          --junit-property github_job_step="Run connectivity test with IPSec (${{ join(matrix.*, ', ') }})"

      - name: Post-test information gathering
        if: ${{ !success() && steps.install-cilium.outcome != 'skipped' }}
        run: |
          kubectl get pods --all-namespaces -o wide
          cilium status
          cilium sysdump --output-filename cilium-sysdump-final-${{ join(matrix.*, '-') }}
        shell: bash {0} # Disable default fail-fast behaviour so that all commands run independently

      - name: Clean up AKS
        if: ${{ always() }}
        run: |
          az group delete --name ${{ env.name }} --yes --no-wait
        shell: bash {0} # Disable default fail-fast behaviour so that all commands run independently

      - name: Upload artifacts
        if: ${{ !success() }}
        uses: actions/upload-artifact@0b2256b8c012f0828dc542b3febcab082c67f72b # v4.3.4
        with:
          name: cilium-sysdumps-${{ matrix.index }}
          path: cilium-sysdump-*.zip

      - name: Upload JUnits [junit]
        if: ${{ always() }}
        uses: actions/upload-artifact@0b2256b8c012f0828dc542b3febcab082c67f72b # v4.3.4
        with:
          name: cilium-junits-${{ matrix.index }}
          path: cilium-junits/*.xml

      - name: Publish Test Results As GitHub Summary
        if: ${{ always() }}
        uses: aanm/junit2md@332ebf0fddd34e91b03a832cfafaa826306558f9 # v0.0.3
        with:
          junit-directory: "cilium-junits"

  merge-upload:
    if: ${{ always() && needs.installation-and-connectivity.result != 'skipped' }}
    name: Merge and Upload Artifacts
    runs-on: ubuntu-latest
    needs: installation-and-connectivity
    steps:
      - name: Merge Sysdumps
        if: ${{ needs.installation-and-connectivity.result == 'failure' }}
        uses: actions/upload-artifact/merge@0b2256b8c012f0828dc542b3febcab082c67f72b # v4.3.4
        with:
          name: cilium-sysdumps
          pattern: cilium-sysdumps-*
          retention-days: 5
          delete-merged: true
        continue-on-error: true
      - name: Merge JUnits
        uses: actions/upload-artifact/merge@0b2256b8c012f0828dc542b3febcab082c67f72b # v4.3.4
        with:
          name: cilium-junits
          pattern: cilium-junits-*
          retention-days: 5
          delete-merged: true

  commit-status-final:
    if: ${{ always() }}
    name: Commit Status Final
    needs: installation-and-connectivity
    runs-on: ubuntu-latest
    steps:
      - name: Set final commit status
        if: ${{ needs.installation-and-connectivity.result != 'skipped' }}
        uses: myrotvorets/set-commit-status-action@3730c0a348a2ace3c110851bed53331bc6406e9f # v2.0.1
        with:
          sha: ${{ inputs.SHA || github.sha }}
          status: ${{ needs.installation-and-connectivity.result }}
      - name: Set final commit status
        if: ${{ needs.installation-and-connectivity.result == 'skipped' }}
        uses: myrotvorets/set-commit-status-action@3730c0a348a2ace3c110851bed53331bc6406e9f # v2.0.1
        with:
          sha: ${{ inputs.SHA || github.sha }}
          status: ${{ github.event_name != 'schedule' && 'success' || 'failure' }}
          description: 'Skipped'