DND-1443: migrate hpa.yaml to autoscaling/v2 and gate it

hpa.yaml emitted apiVersion: autoscaling/v2beta1, removed in Kubernetes 1.25.
On any cluster >= 1.25, autoscaling.enabled=true produced a
HorizontalPodAutoscaler the API server rejects; kubeconform (strict) also
failed it ("could not find schema for HorizontalPodAutoscaler").

- hpa.yaml: apiVersion autoscaling/v2beta1 -> autoscaling/v2, and the metrics
  shape migrated to the v2 form (spec.metrics[].resource.target.type=Utilization
  + averageUtilization, instead of targetAverageUtilization).
- Add test-values/autoscaling.yaml (HPA enabled, CPU + memory targets) and
  register it in the lint-render matrix so kubeconform now validates the HPA.
  Previously no gated scenario enabled autoscaling, so the HPA was never
  rendered or checked. helm-diff globs test-values/*.yaml and picks it up too.

No Chart.yaml version bump (deferred, handled separately), so verify-version
fails by design for now.

Verified: renders autoscaling/v2 with the v2 metrics shape; kubeconform -strict
(no HPA skip, CI-equivalent) validates all rendered resources; helm lint +
actionlint clean.

Author: CRThaze

.github/workflows/lint-render.yaml

@@ -35,6 +35,7 @@ jobs:
           - test-values/rackspace.yaml
           - test-values/multi-backend.yaml
           - test-values/ingress.yaml
+          - test-values/autoscaling.yaml
     steps:
       - name: Checkout
         uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2

charts/s3proxy/templates/hpa.yaml

@@ -1,5 +1,5 @@
 {{- if .Values.autoscaling.enabled }}
-apiVersion: autoscaling/v2beta1
+apiVersion: autoscaling/v2
 kind: HorizontalPodAutoscaler
 metadata:
   name: {{ include "s3proxy.fullname" . }}
@@ -17,12 +17,16 @@ spec:
     - type: Resource
       resource:
         name: cpu
-        targetAverageUtilization: {{ .Values.autoscaling.targetCPUUtilizationPercentage }}
+        target:
+          type: Utilization
+          averageUtilization: {{ .Values.autoscaling.targetCPUUtilizationPercentage }}
   {{- end }}
   {{- if .Values.autoscaling.targetMemoryUtilizationPercentage }}
     - type: Resource
       resource:
         name: memory
-        targetAverageUtilization: {{ .Values.autoscaling.targetMemoryUtilizationPercentage }}
+        target:
+          type: Utilization
+          averageUtilization: {{ .Values.autoscaling.targetMemoryUtilizationPercentage }}
   {{- end }}
 {{- end }}

test-values/autoscaling.yaml

@@ -0,0 +1,25 @@
+# Render/lint scenario: HorizontalPodAutoscaler enabled, exercising both the CPU
+# and memory target metrics. kubeconform validates the rendered HPA against the
+# Kubernetes schema, so this scenario is the gate that keeps hpa.yaml on the
+# supported autoscaling/v2 API (the removed autoscaling/v2beta1 was rejected by
+# kubeconform on k8s >= 1.25 — DND-1443).
+config:
+  auth:
+    type: aws-v4
+    identity: test-access-key
+    secret: test-secret-key
+  backends:
+    filesystem:
+      enabled: true
+autoscaling:
+  enabled: true
+  minReplicas: 2
+  maxReplicas: 5
+  targetCPUUtilizationPercentage: 80
+  targetMemoryUtilizationPercentage: 75
+persistence:
+  enabled: false
+resources:
+  requests:
+    cpu: 100m
+    memory: 256Mi