Enhance debugging output for failed service boot and test failures in CI workflow

Author: emmanuelmathot

.github/workflows/helm-tests.yml

@@ -118,7 +118,40 @@ jobs:
       - name: cleanup if services fail to boot
         if: steps.watchservices.outcome == 'failure'
         run: |
-          echo "The previous step failed or timed out."
+          echo "The watchservices step failed or timed out. Extracting pod logs for debugging..."
+          
+          # Get and display all pods status
+          echo "===== Pod Status ====="
+          kubectl get pods
+          
+          # Extract logs from database pod
+          echo "===== Database Pod Logs ====="
+          kubectl get pod | grep "^db-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} --tail=100 || echo "Could not get database logs"
+          
+          # Extract logs from pgstacbootstrap pod
+          echo "===== PGSTACBootstrap Pod Logs ====="
+          kubectl get pod | grep "^pgstacbootstrap-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} --tail=100 || echo "Could not get pgstacbootstrap logs"
+          
+          # Extract logs from raster pod init container (wait-for-pgstacbootstrap)
+          echo "===== Raster Pod Init Container Logs (wait-for-pgstacbootstrap) ====="
+          kubectl get pod | grep "^raster-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} -c wait-for-pgstacbootstrap --tail=100 || echo "Could not get raster init container logs"
+          
+          # Extract logs from raster pod main container
+          echo "===== Raster Pod Main Container Logs ====="
+          kubectl get pod | grep "^raster-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} --tail=100 || echo "Could not get raster main container logs"
+          
+          # Extract logs from vector pod
+          echo "===== Vector Pod Logs ====="
+          kubectl get pod | grep "^vector-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} --tail=100 || echo "Could not get vector logs"
+          
+          # Extract logs from stac pod
+          echo "===== STAC Pod Logs ====="
+          kubectl get pod | grep "^stac-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} --tail=100 || echo "Could not get STAC logs"
+          
+          # Check if pods are in pending state or have issues
+          echo "===== Pod Descriptions for Troubleshooting ====="
+          kubectl get pod | grep "$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl describe pod/{} || echo "Could not describe pods"
+          
           # force GH action to show failed result
           exit 128
 
@@ -145,18 +178,51 @@ jobs:
           echo $RASTER_ENDPOINT
           echo '#################################'
 
-          pytest .github/workflows/tests/test_vector.py || kubectl logs svc/vector-$RELEASE_NAME
-          pytest .github/workflows/tests/test_stac.py
+          pytest .github/workflows/tests/test_vector.py || kubectl logs svc/vector
+          pytest .github/workflows/tests/test_stac.py || kubectl logs svc/stac
           # TODO: fix raster tests
-          #pytest .github/workflows/tests/test_raster.py
+          #pytest .github/workflows/tests/test_raster.py || kubectl logs svc/raster
 
       - name: error if tests failed
         if: steps.testrunner.outcome == 'failure'
         run: |
-          echo "The previous step failed or timed out."
+          echo "The tests failed. Extracting pod logs for debugging..."
+          
+          # Get and display all pods status
+          echo "===== Pod Status ====="
+          kubectl get pods
+          
+          # Extract logs from database pod
+          echo "===== Database Pod Logs ====="
+          kubectl get pod | grep "^db-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} --tail=100 || echo "Could not get database logs"
+          
+          # Extract logs from pgstacbootstrap pod
+          echo "===== PGSTACBootstrap Pod Logs ====="
+          kubectl get pod | grep "^pgstacbootstrap-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} --tail=100 || echo "Could not get pgstacbootstrap logs"
+          
+          # Extract logs from raster pod init container (wait-for-pgstacbootstrap)
+          echo "===== Raster Pod Init Container Logs (wait-for-pgstacbootstrap) ====="
+          kubectl get pod | grep "^raster-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} -c wait-for-pgstacbootstrap --tail=100 || echo "Could not get raster init container logs"
+          
+          # Extract logs from raster pod main container
+          echo "===== Raster Pod Main Container Logs ====="
+          kubectl get pod | grep "^raster-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} --tail=100 || echo "Could not get raster main container logs"
+          
+          # Extract logs from vector pod
+          echo "===== Vector Pod Logs ====="
+          kubectl get pod | grep "^vector-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} --tail=100 || echo "Could not get vector logs"
+          
+          # Extract logs from stac pod
+          echo "===== STAC Pod Logs ====="
+          kubectl get pod | grep "^stac-$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl logs pod/{} --tail=100 || echo "Could not get STAC logs"
+          
+          # Check if pods are in pending state or have issues
+          echo "===== Pod Descriptions for Troubleshooting ====="
+          kubectl get pod | grep "$RELEASE_NAME" | cut -d' ' -f1 | xargs -I{} kubectl describe pod/{} || echo "Could not describe pods"
+          
           # force GH action to show failed result
           exit 128
 
       - name: helm uninstall eoapi templates
         run: |
-          helm uninstall $RELEASE_NAME
+          helm uninstall $RELEASE_NAME