chore: add debugging

Author: mchmarny

.github/workflows/integration-gcp.yml

@@ -130,21 +130,25 @@ jobs:
           GCP_ZONE: "${{ env.TF_VAR_zone }}"
           GCP_SERVICE_ACCOUNT: "${{ env.SERVICE_ACCOUNT }}"
           NVSENTINEL_VERSION: "${{ steps.ref-name.outputs.value }}"
-        run: tests/uat/install-apps.sh
+        run: |
+          set -euxo pipefail
+          tests/uat/install-apps.sh
 
       # Test
       - name: Run UAT Tests
         id: tests
         if: steps.apps.outcome == 'success'
         shell: bash
-        run: tests/uat/tests.sh
+        run: |
+          set -euxo pipefail
+          tests/uat/tests.sh
 
       # Teardown
       - name: Destroy Cluster
         if: always() && steps.cluster.outcome != 'skipped' && env.SKIP_DELETE != 'true'
         shell: bash
         run: |
-          set -euo pipefail
+          set -euxo pipefail
           cd tests/uat/gcp/cluster
           terraform destroy -auto-approve
 

tests/uat/gcp/gpu-operator-values.yaml

@@ -34,4 +34,17 @@ toolkit:
   enabled: true
   env:
   - name: RUNTIME_CONFIG_SOURCE
-    value: file
+    value: file
+
+# Fix deprecated node-role label warnings
+node-feature-discovery:
+  master:
+    affinity:
+      nodeAffinity:
+        preferredDuringSchedulingIgnoredDuringExecution:
+        - weight: 1
+          preference:
+            matchExpressions:
+            - key: "node-role.kubernetes.io/control-plane"
+              operator: In
+              values: [""]

tests/uat/install-apps.sh

@@ -59,6 +59,32 @@ GCP_SERVICE_ACCOUNT="${GCP_SERVICE_ACCOUNT:-}"
 # ARM64-specific values file (if needed)
 NVSENTINEL_ARM64_VALUES="${REPO_ROOT}/distros/kubernetes/nvsentinel/values-tilt-arm64.yaml"
 
+
+# Print out variables for debugging (alphabetical order)
+log "Using configuration (raw):"
+log "  - AWS_REGION: $AWS_REGION"
+log "  - CERT_MANAGER_VALUES: $CERT_MANAGER_VALUES"
+log "  - CERT_MANAGER_VERSION: $CERT_MANAGER_VERSION"
+log "  - CLUSTER_NAME: $CLUSTER_NAME"
+log "  - CSP: $CSP"
+log "  - FAKE_GPU_NODE_COUNT: $FAKE_GPU_NODE_COUNT" 
+log "  - GCP_PROJECT_ID: $GCP_PROJECT_ID"
+log "  - GCP_SERVICE_ACCOUNT: $GCP_SERVICE_ACCOUNT"
+log "  - GCP_ZONE: $GCP_ZONE"
+log "  - GPU_OPERATOR_VALUES: $GPU_OPERATOR_VALUES"
+log "  - GPU_OPERATOR_VERSION: $GPU_OPERATOR_VERSION"
+log "  - KWOK_VERSION: $KWOK_VERSION (chart: $KWOK_CHART_VERSION)"
+log "  - NVSENTINEL_ARM64_VALUES: $NVSENTINEL_ARM64_VALUES"
+log "  - NVSENTINEL_CHART: $NVSENTINEL_CHART"
+log "  - NVSENTINEL_VALUES: $NVSENTINEL_VALUES"
+log "  - NVSENTINEL_VERSION: $NVSENTINEL_VERSION"
+log "  - PROMETHEUS_OPERATOR_VERSION: $PROMETHEUS_OPERATOR_VERSION"
+log "  - PROMETHEUS_VALUES: $PROMETHEUS_VALUES"
+log "  - RESOURCE_QUOTA_RESOURCE: $RESOURCE_QUOTA_RESOURCE"
+log "  - VALUES_DIR: $VALUES_DIR"
+log ""
+
+
 install_prometheus_operator() {
     log "Installing Prometheus Operator (version $PROMETHEUS_OPERATOR_VERSION)..."