fix: apply CodeRabbit auto-fixes

Fixed 4 file(s) based on 5 unresolved review comments.

Co-authored-by: CodeRabbit <noreply@coderabbit.ai>

Author: coderabbitai[bot]

config/prometheus-adapter/custom-metrics-config.yaml

@@ -9,7 +9,7 @@
 #   kubectl rollout restart deployment/prometheus-adapter -n monitoring
 #
 # Verify metrics are registered:
-#   kubectl get --raw /apis/custom.metrics.k8s.io/v1beta1 | jq .
+#   kubectl get --raw /apis/external.metrics.k8s.io/v1beta1 | jq .
 #
 apiVersion: v1
 kind: ConfigMap

config/prometheus-adapter/kustomization.yaml

@@ -14,8 +14,28 @@ namespace: monitoring
 resources:
   - custom-metrics-config.yaml
 
-# Patch removed: The prometheus-adapter Deployment is not included in this
-# kustomization's resources, so the patch has no valid target. Users should
-# apply this ConfigMap to their cluster and manually configure their existing
-# prometheus-adapter Deployment to mount it, or include the prometheus-adapter
-# base resources in this kustomization before re-adding the patch.
+# To apply this configuration, either:
+# 1. Include the prometheus-adapter Deployment base in resources above, then
+#    uncomment the patches section below, OR
+# 2. Manually configure your existing prometheus-adapter Deployment to mount
+#    this ConfigMap at /etc/adapter/config.yaml and pass --config=/etc/adapter/config.yaml
+#
+# patches:
+#   - target:
+#       kind: Deployment
+#       name: prometheus-adapter
+#     patch: |-
+#       - op: add
+#         path: /spec/template/spec/volumes/-
+#         value:
+#           name: adapter-config
+#           configMap:
+#             name: agentrax-custom-metrics
+#       - op: add
+#         path: /spec/template/spec/containers/0/volumeMounts/-
+#         value:
+#           name: adapter-config
+#           mountPath: /etc/adapter
+#       - op: add
+#         path: /spec/template/spec/containers/0/args/-
+#         value: --config=/etc/adapter/config.yaml

internal/controller/agentdeployment_controller_test.go

@@ -18,7 +18,6 @@ package controller
 
 import (
 	"context"
-	"fmt"
 	"time"
 
 	. "github.com/onsi/ginkgo/v2"
@@ -142,12 +141,10 @@ func deleteChildResources(key types.NamespacedName) {
 	hpa := &autoscalingv2.HorizontalPodAutoscaler{}
 	err := k8sClient.Get(ctx, key, hpa)
 	if err != nil && !apierrors.IsNotFound(err) {
-		panic(fmt.Sprintf("unexpected error reading HPA during cleanup: %v", err))
+		Expect(err).NotTo(HaveOccurred(), "unexpected error reading HPA during cleanup")
 	}
 	if err == nil {
-		if err := k8sClient.Delete(ctx, hpa); err != nil {
-			panic(fmt.Sprintf("failed to delete HPA during cleanup: %v", err))
-		}
+		Expect(k8sClient.Delete(ctx, hpa)).To(Succeed(), "HPA cleanup delete should succeed")
 		Eventually(func() bool {
 			return apierrors.IsNotFound(k8sClient.Get(ctx, key, &autoscalingv2.HorizontalPodAutoscaler{}))
 		}, testTimeout, testInterval).Should(BeTrue(), "child HPA should be deleted")

internal/quota/enforcer.go

@@ -254,28 +254,37 @@ func evalQuotaRules(
 	isUpdate bool,
 	prevMaxReplicas int32, // 0 for creates; used by per-agent ceiling check
 ) (bool, string) {
-	projAgents := committedUsage.UsedAgents + inFlight.agents + delta.agents
-	projGPUs := committedUsage.UsedGPUs + inFlight.gpus + delta.gpus
-	projReplicas := committedUsage.UsedTotalReplicas + inFlight.replicas + delta.replicas
+	// Reject the math.MaxInt32 sentinel used by gpusForAD to signal overflow.
+	// If delta.gpus is MaxInt32, the GPU calculation overflowed and we must
+	// fail closed to prevent undercount-based admission.
+	if delta.gpus == math.MaxInt32 {
+		return false, "GPU resource calculation overflowed; request denied"
+	}
+
+	// Perform projection calculations in int64 to prevent overflow, then
+	// compare against int64-converted quota limits.
+	projAgents := int64(committedUsage.UsedAgents) + int64(inFlight.agents) + int64(delta.agents)
+	projGPUs := int64(committedUsage.UsedGPUs) + int64(inFlight.gpus) + int64(delta.gpus)
+	projReplicas := int64(committedUsage.UsedTotalReplicas) + int64(inFlight.replicas) + int64(delta.replicas)
 
 	// For UPDATE requests, only reject when the delta increases a dimension that
 	// is already at or over quota. If quota was lowered below current usage, the
 	// existing ADs are already OverQuota (indicated by the TQ condition) — we
 	// must not block updates that don't make things worse, otherwise finalizer
 	// removal and spec corrections are deadlocked.
-	if projAgents > quota.MaxAgents && (!isUpdate || delta.agents > 0) {
+	if projAgents > int64(quota.MaxAgents) && (!isUpdate || delta.agents > 0) {
 		return false, fmt.Sprintf(
 			"would exceed maxAgents (%d): current=%d in-flight=%d delta=%d",
 			quota.MaxAgents, committedUsage.UsedAgents, inFlight.agents, delta.agents,
 		)
 	}
-	if projGPUs > quota.MaxGPUs && (!isUpdate || delta.gpus > 0) {
+	if projGPUs > int64(quota.MaxGPUs) && (!isUpdate || delta.gpus > 0) {
 		return false, fmt.Sprintf(
 			"would exceed maxGPUs (%d): current=%d in-flight=%d delta=%d",
 			quota.MaxGPUs, committedUsage.UsedGPUs, inFlight.gpus, delta.gpus,
 		)
 	}
-	if projReplicas > quota.MaxTotalReplicas && (!isUpdate || delta.replicas > 0) {
+	if projReplicas > int64(quota.MaxTotalReplicas) && (!isUpdate || delta.replicas > 0) {
 		return false, fmt.Sprintf(
 			"would exceed maxTotalReplicas (%d): current=%d in-flight=%d delta=%d",
 			quota.MaxTotalReplicas, committedUsage.UsedTotalReplicas, inFlight.replicas, delta.replicas,