use latest release for initial CI cluster setup

Author: sjpb

.github/workflows/stackhpc.yml

@@ -44,7 +44,10 @@ jobs:
       CI_CLOUD: ${{ vars.CI_CLOUD }} # default from repo settings
       TF_VAR_os_version: ${{ matrix.os_version }}
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          fetch-tags: true
 
       - name: Override CI_CLOUD if PR label is present
         if: ${{ github.event_name == 'pull_request' }}
@@ -76,6 +79,14 @@ jobs:
         run: cat environments/.stackhpc/bastion_fingerprints >> ~/.ssh/known_hosts
         shell: bash
 
+      - name: Find the latest release
+        run: |
+          echo LATEST_RELEASE_TAG=$(curl -s https://api.github.com/repos/stackhpc/ansible-slurm-appliance/releases/latest | jq -r .tag_name) >> "$GITHUB_ENV"
+          echo LATEST_RELEASE_TAG: $LATEST_RELEASE_TAG
+      
+      - name: Checkout latest release
+        run: git checkout $LATEST_RELEASE_TAG
+      
       - name: Install ansible etc
         run: dev/setup-env.sh
 
@@ -103,7 +114,7 @@ jobs:
         env:
           DEMO_USER_PASSWORD: ${{ secrets.TEST_USER_PASSWORD }}
 
-      - name: Provision nodes using fat image
+      - name: Provision nodes using latest release image
         id: provision_servers
         run: |
           . venv/bin/activate
@@ -119,15 +130,15 @@ jobs:
           tofu destroy -auto-approve -var-file="${{ env.CI_CLOUD }}.tfvars"
         if: failure() && steps.provision_servers.outcome == 'failure'
 
-      - name: Configure cluster
+      - name: Configure cluster at latest release
         run: |
           . venv/bin/activate
           . environments/.stackhpc/activate
           ansible all -m wait_for_connection
           ansible-playbook -v ansible/site.yml
           ansible-playbook -v ansible/ci/check_slurm.yml
 
-      - name: Run MPI-based tests
+      - name: Run MPI-based tests at latest release
         run: |
           . venv/bin/activate
           . environments/.stackhpc/activate
@@ -170,23 +181,33 @@ jobs:
         env:
           DEMO_USER_PASSWORD: ${{ secrets.TEST_USER_PASSWORD }}
 
-      - name: Test reimage of login and control nodes (via rebuild adhoc)
+      - name: Switch to current branch
+        run: git checkout -
+
+      - name: Reimage login and control nodes to image in current branch
+        id: reimage_non_compute
+        run: |
+          . venv/bin/activate
+          . environments/.stackhpc/activate
+          cd $APPLIANCES_ENVIRONMENT_ROOT/tofu
+          tofu apply -auto-approve -var-file="${{ env.CI_CLOUD }}.tfvars" -var-file=cluster_image.latest.tfvars.json
+          
+      - name: Configure cluster using current branch
         run: |
           . venv/bin/activate
           . environments/.stackhpc/activate
-          ansible-playbook -v --limit control,login ansible/adhoc/rebuild.yml
+          ansible all -m wait_for_connection
           ansible-playbook -v ansible/site.yml
           ansible-playbook -v ansible/ci/check_slurm.yml
 
-      - name: Test compute node reboot and compute-init
+      - name: Reimage compute nodes to image in current branch using slurm - tests compute-init
         run: |
           . venv/bin/activate
           . environments/.stackhpc/activate
-          ansible-playbook -v --limit compute ansible/adhoc/rebuild.yml
-          ansible-playbook -v ansible/ci/check_slurm.yml
           ansible-playbook -v ansible/adhoc/reboot_via_slurm.yml
+          ansible-playbook -v ansible/ci/check_slurm.yml
 
-      - name: Check sacct state survived reimage
+      - name: Check sacct state survived reimage to current branch
         run: |
           . venv/bin/activate
           . environments/.stackhpc/activate
@@ -198,16 +219,16 @@ jobs:
           . environments/.stackhpc/activate
           ansible-playbook -vv ansible/ci/check_grafana.yml
 
+      - name: Run MPI-based tests again in current branch
+        run: |
+          . venv/bin/activate
+          . environments/.stackhpc/activate
+          ansible-playbook -vv ansible/adhoc/hpctests.yml
+
       - name: Delete infrastructure
         run: |
           . venv/bin/activate
           . environments/.stackhpc/activate
           cd $APPLIANCES_ENVIRONMENT_ROOT/tofu
           tofu destroy -auto-approve -var-file="${{ env.CI_CLOUD }}.tfvars"
         if: ${{ success() || cancelled() }}
-
-      # - name: Delete images
-      #   run: |
-      #     . venv/bin/activate
-      #     . environments/.stackhpc/activate
-      #     ansible-playbook -vv ansible/ci/delete_images.yml