Feedback from PR review.

Author: pantierra

.github/workflows/ci.yml

@@ -55,7 +55,7 @@ jobs:
 
       - name: Install Python dependencies
         run: |
-          pip install pytest pytest-timeout requests
+          pip install pytest pytest-timeout requests psycopg2-binary
 
       - name: Start K3s cluster
         uses: jupyterhub/action-k3s-helm@v4
@@ -114,11 +114,6 @@ jobs:
           echo "=== Post-deployment validation ==="
           ./scripts/test.sh check-deployment
 
-      - name: Install Python dependencies
-        run: |
-          python -m pip install --upgrade pip
-          pip install pytest requests psycopg2-binary
-
       - name: Run integration tests
         run: |
           export RELEASE_NAME="$RELEASE_NAME"

.github/workflows/tests/test_autoscaling.py

@@ -11,18 +11,14 @@
 
 
 def get_namespace():
-    """Get the target namespace from environment or default."""
     return os.environ.get("NAMESPACE", "eoapi")
 
 
 def get_release_name():
-    """Get the release name from environment or default."""
     return os.environ.get("RELEASE_NAME", "eoapi")
 
 
 def get_base_url():
-    """Get the base URL for API endpoints."""
-    # Try to detect ingress or use port-forward
     namespace = get_namespace()
 
     # Check if we have an ingress
@@ -49,12 +45,10 @@ def get_base_url():
                 except requests.RequestException:
                     pass
 
-    # Fallback to localhost (assuming port-forward)
     return "http://localhost:8080"
 
 
 def kubectl_get(resource, namespace=None, label_selector=None, output="json"):
-    """Execute kubectl get command with optional parameters."""
     cmd = ["kubectl", "get", resource]
 
     if namespace:
@@ -71,7 +65,6 @@ def kubectl_get(resource, namespace=None, label_selector=None, output="json"):
 
 
 def get_pod_metrics(namespace, service_name):
-    """Get current CPU and memory metrics for service pods."""
     release_name = get_release_name()
     result = subprocess.run(
         [
@@ -105,7 +98,6 @@ def get_pod_metrics(namespace, service_name):
 
 
 def get_hpa_status(namespace, hpa_name):
-    """Get current HPA status and metrics."""
     result = kubectl_get("hpa", namespace=namespace, output="json")
     if result.returncode != 0:
         return None
@@ -119,7 +111,6 @@ def get_hpa_status(namespace, hpa_name):
 
 
 def get_pod_count(namespace, service_name):
-    """Get current number of running pods for a service."""
     release_name = get_release_name()
     result = kubectl_get(
         "pods",

docs/examples/values-autoscaling.yaml

@@ -1,4 +1,23 @@
 # Example values for eoAPI with core monitoring and autoscaling enabled
+#
+# To use this configuration:
+#
+# 1. Update the ingress.host to your actual domain
+# 2. Adjust scaling targets based on your load testing results
+# 3. Monitor resource usage and adjust requests/limits accordingly
+# 4. Consider enabling TLS for production deployments
+#
+# IMPORTANT: This configuration enables monitoring components that are
+# disabled by default. This is required for autoscaling to work.
+#
+# For observability and dashboards, install the separate eoapi-observability chart:
+# helm install eoapi-obs eoapi/eoapi-observability --namespace eoapi
+#
+# Load testing recommendations:
+# - Test each service endpoint individually
+# - Monitor HPA metrics: kubectl get hpa -n eoapi -w
+# - Check custom metrics: kubectl get --raw "/apis/custom.metrics.k8s.io/v1beta1"
+# - Review Prometheus targets to ensure metrics collection is working
 
 gitSha: "latest"
 
@@ -187,26 +206,3 @@ pgstacBootstrap:
       limits:
         cpu: "1024m"
         memory: "2048Mi"
-
-######################
-# ADDITIONAL NOTES
-######################
-#
-# To use this configuration:
-#
-# 1. Update the ingress.host to your actual domain
-# 2. Adjust scaling targets based on your load testing results
-# 3. Monitor resource usage and adjust requests/limits accordingly
-# 4. Consider enabling TLS for production deployments
-#
-# IMPORTANT: This configuration enables monitoring components that are
-# disabled by default. This is required for autoscaling to work.
-#
-# For observability and dashboards, install the separate eoapi-observability chart:
-# helm install eoapi-obs eoapi/eoapi-observability --namespace eoapi
-#
-# Load testing recommendations:
-# - Test each service endpoint individually
-# - Monitor HPA metrics: kubectl get hpa -n eoapi -w
-# - Check custom metrics: kubectl get --raw "/apis/custom.metrics.k8s.io/v1beta1"
-# - Review Prometheus targets to ensure metrics collection is working