From 36408838aeb791c8731dece68e3115e6f5565e7e Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Thu, 25 Jun 2026 12:15:25 +0100
Subject: [PATCH 01/14] feat(otel): add Karpenter integration test and
 Terraform setup

---
 terraform/eks/daemon/otel/main.tf             | 223 +++++++++++++++++-
 terraform/eks/daemon/otel/variables.tf        |   5 +
 .../integrations/karpenter/karpenter_test.go  | 117 +++++++++
 .../integrations/karpenter/metrics_test.go    |  63 +++++
 .../otel/integrations/karpenter/setup_test.go | 101 ++++++++
 util/otelmetrics/source_registry.go           |   1 +
 6 files changed, 508 insertions(+), 2 deletions(-)
 create mode 100644 test/otel/integrations/karpenter/karpenter_test.go
 create mode 100644 test/otel/integrations/karpenter/metrics_test.go
 create mode 100644 test/otel/integrations/karpenter/setup_test.go

diff --git a/terraform/eks/daemon/otel/main.tf b/terraform/eks/daemon/otel/main.tf
index 22b3d5de2..5a1097cd6 100644
--- a/terraform/eks/daemon/otel/main.tf
+++ b/terraform/eks/daemon/otel/main.tf
@@ -13,11 +13,14 @@ module "basic_components" {
 }
 
 locals {
-  aws_eks = "aws eks --region ${var.region}"
+  aws_eks      = "aws eks --region ${var.region}"
+  cluster_name = "cwagent-eks-integ-${module.common.testing_id}"
 }
 
+data "aws_partition" "current" {}
+
 resource "aws_eks_cluster" "this" {
-  name     = "cwagent-eks-integ-${module.common.testing_id}"
+  name     = local.cluster_name
   role_arn = module.basic_components.role_arn
   version  = var.k8s_version
   vpc_config {
@@ -110,6 +113,130 @@ resource "aws_eks_addon" "pod_identity_agent" {
   addon_name   = "eks-pod-identity-agent"
 }
 
+# -----------------------------------------------------------------------------
+# Karpenter Controller IAM Role (Pod Identity)
+# -----------------------------------------------------------------------------
+
+resource "aws_iam_role" "karpenter_controller" {
+  name = "cwagent-otel-karpenter-${module.common.testing_id}"
+  assume_role_policy = jsonencode({
+    Version = "2012-10-17"
+    Statement = [{
+      Effect    = "Allow"
+      Principal = { Service = "pods.eks.amazonaws.com" }
+      Action    = ["sts:AssumeRole", "sts:TagSession"]
+    }]
+  })
+}
+
+resource "aws_iam_role_policy" "karpenter_controller" {
+  name = "KarpenterControllerPolicy"
+  role = aws_iam_role.karpenter_controller.id
+  policy = jsonencode({
+    Version = "2012-10-17"
+    Statement = [
+      {
+        Sid    = "EC2Permissions"
+        Effect = "Allow"
+        Action = [
+          "ec2:CreateFleet",
+          "ec2:CreateLaunchTemplate",
+          "ec2:CreateTags",
+          "ec2:DeleteLaunchTemplate",
+          "ec2:DescribeAvailabilityZones",
+          "ec2:DescribeImages",
+          "ec2:DescribeInstances",
+          "ec2:DescribeInstanceTypeOfferings",
+          "ec2:DescribeInstanceTypes",
+          "ec2:DescribeLaunchTemplates",
+          "ec2:DescribeSecurityGroups",
+          "ec2:DescribeSubnets",
+          "ec2:RunInstances",
+          "ec2:TerminateInstances",
+        ]
+        Resource = "*"
+      },
+      {
+        Sid      = "PassRoleToEC2"
+        Effect   = "Allow"
+        Action   = "iam:PassRole"
+        Resource = aws_iam_role.node_role.arn
+      },
+      {
+        Sid      = "EKSPermissions"
+        Effect   = "Allow"
+        Action   = "eks:DescribeCluster"
+        Resource = aws_eks_cluster.this.arn
+      },
+      {
+        Sid      = "SSMGetParameter"
+        Effect   = "Allow"
+        Action   = "ssm:GetParameter"
+        Resource = "arn:${data.aws_partition.current.partition}:ssm:${var.region}::parameter/aws/service/*"
+      },
+      {
+        Sid      = "PricingPermissions"
+        Effect   = "Allow"
+        Action   = "pricing:GetProducts"
+        Resource = "*"
+      },
+      {
+        Sid    = "SQSPermissions"
+        Effect = "Allow"
+        Action = [
+          "sqs:DeleteMessage",
+          "sqs:GetQueueUrl",
+          "sqs:ReceiveMessage",
+        ]
+        Resource = aws_sqs_queue.karpenter_interruption.arn
+      },
+    ]
+  })
+}
+
+# Karpenter Interruption Queue (SQS)
+resource "aws_sqs_queue" "karpenter_interruption" {
+  name                      = "cwagent-otel-karpenter-${module.common.testing_id}"
+  message_retention_seconds = 300
+  sqs_managed_sse_enabled   = true
+}
+
+resource "aws_sqs_queue_policy" "karpenter_interruption" {
+  queue_url = aws_sqs_queue.karpenter_interruption.id
+  policy = jsonencode({
+    Version = "2012-10-17"
+    Statement = [{
+      Effect    = "Allow"
+      Principal = { Service = ["events.amazonaws.com", "sqs.amazonaws.com"] }
+      Action    = "sqs:SendMessage"
+      Resource  = aws_sqs_queue.karpenter_interruption.arn
+    }]
+  })
+}
+
+# Karpenter Pod Identity Association
+resource "aws_eks_pod_identity_association" "karpenter_controller" {
+  depends_on      = [aws_eks_addon.pod_identity_agent]
+  cluster_name    = aws_eks_cluster.this.name
+  namespace       = "kube-system"
+  service_account = "karpenter"
+  role_arn        = aws_iam_role.karpenter_controller.arn
+}
+
+# Tag subnets and security group for Karpenter discovery
+resource "aws_ec2_tag" "subnet_karpenter" {
+  count       = length(module.basic_components.public_subnet_ids)
+  resource_id = module.basic_components.public_subnet_ids[count.index]
+  key         = "karpenter.sh/discovery"
+  value       = aws_eks_cluster.this.name
+}
+
+resource "aws_ec2_tag" "sg_karpenter" {
+  resource_id = module.basic_components.security_group
+  key         = "karpenter.sh/discovery"
+  value       = aws_eks_cluster.this.name
+}
+
 # --- Update kubeconfig ---
 
 resource "null_resource" "kubectl" {
@@ -119,6 +246,88 @@ resource "null_resource" "kubectl" {
   }
 }
 
+# --- Karpenter Helm install ---
+
+resource "helm_release" "karpenter" {
+  name             = "karpenter"
+  repository       = "oci://public.ecr.aws/karpenter"
+  chart            = "karpenter"
+  version          = var.karpenter_version
+  namespace        = "kube-system"
+  create_namespace = false
+
+  set = [
+    { name = "settings.clusterName", value = aws_eks_cluster.this.name },
+    { name = "settings.clusterEndpoint", value = aws_eks_cluster.this.endpoint },
+    { name = "settings.interruptionQueue", value = aws_sqs_queue.karpenter_interruption.name },
+    { name = "controller.resources.requests.cpu", value = "100m" },
+    { name = "controller.resources.requests.memory", value = "256Mi" },
+  ]
+
+  depends_on = [
+    aws_eks_addon.pod_identity_agent,
+    aws_eks_pod_identity_association.karpenter_controller,
+    null_resource.kubectl,
+  ]
+}
+
+# --- Karpenter NodePool + EC2NodeClass (minimal, for metric generation) ---
+
+resource "null_resource" "karpenter_nodepool" {
+  depends_on = [helm_release.karpenter]
+  provisioner "local-exec" {
+    command = <<-EOT
+      cat <<'EOF' | kubectl apply -f -
+      apiVersion: karpenter.sh/v1
+      kind: NodePool
+      metadata:
+        name: default
+      spec:
+        template:
+          spec:
+            nodeClassRef:
+              group: karpenter.k8s.aws
+              kind: EC2NodeClass
+              name: default
+            requirements:
+              - key: "karpenter.k8s.aws/instance-category"
+                operator: In
+                values: ["t"]
+              - key: "karpenter.k8s.aws/instance-size"
+                operator: In
+                values: ["medium", "large"]
+              - key: "kubernetes.io/arch"
+                operator: In
+                values: ["amd64"]
+              - key: "karpenter.sh/capacity-type"
+                operator: In
+                values: ["on-demand"]
+        limits:
+          cpu: "4"
+          memory: "8Gi"
+        disruption:
+          consolidationPolicy: WhenEmpty
+          consolidateAfter: 60s
+      ---
+      apiVersion: karpenter.k8s.aws/v1
+      kind: EC2NodeClass
+      metadata:
+        name: default
+      spec:
+        role: "${aws_iam_role.node_role.name}"
+        amiSelectorTerms:
+          - alias: al2023@latest
+        subnetSelectorTerms:
+          - tags:
+              karpenter.sh/discovery: "${aws_eks_cluster.this.name}"
+        securityGroupSelectorTerms:
+          - tags:
+              karpenter.sh/discovery: "${aws_eks_cluster.this.name}"
+      EOF
+    EOT
+  }
+}
+
 # --- Helm chart install ---
 
 data "external" "clone_helm_chart" {
@@ -140,12 +349,14 @@ resource "helm_release" "aws_observability" {
     { name = "clusterName", value = aws_eks_cluster.this.name },
     { name = "region", value = var.region },
     { name = "otelContainerInsights.enabled", value = "true" },
+    { name = "otelContainerInsights.metrics.karpenter.enabled", value = "true" },
   ]
 
   depends_on = [
     aws_eks_addon.pod_identity_agent,
     null_resource.kubectl,
     data.external.clone_helm_chart,
+    helm_release.karpenter,
   ]
 }
 
@@ -351,6 +562,7 @@ resource "null_resource" "validator" {
   depends_on = [
     null_resource.restart_pods,
     kubernetes_deployment_v1.nginx_test,
+    null_resource.karpenter_nodepool,
   ]
 
   triggers = { always_run = timestamp() }
@@ -368,6 +580,13 @@ resource "null_resource" "validator" {
         -computeType=EKS \
         -eksDeploymentStrategy=DAEMON \
         -region=${var.region}
+
+      echo "Running OTEL integrations tests (Karpenter)..."
+      go test -tags integration -timeout 1h -v ./test/otel/integrations/karpenter/... \
+        -eksClusterName=${aws_eks_cluster.this.name} \
+        -computeType=EKS \
+        -eksDeploymentStrategy=DAEMON \
+        -region=${var.region}
     EOT
   }
 }
diff --git a/terraform/eks/daemon/otel/variables.tf b/terraform/eks/daemon/otel/variables.tf
index 9149462f2..7ed580859 100644
--- a/terraform/eks/daemon/otel/variables.tf
+++ b/terraform/eks/daemon/otel/variables.tf
@@ -40,3 +40,8 @@ variable "instance_type" {
   type    = string
   default = "t3.medium"
 }
+
+variable "karpenter_version" {
+  type    = string
+  default = "1.4.0"
+}
diff --git a/test/otel/integrations/karpenter/karpenter_test.go b/test/otel/integrations/karpenter/karpenter_test.go
new file mode 100644
index 000000000..b3e86b52e
--- /dev/null
+++ b/test/otel/integrations/karpenter/karpenter_test.go
@@ -0,0 +1,117 @@
+//go:build integration
+
+// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
+// SPDX-License-Identifier: MIT
+
+package karpenter
+
+import (
+	"context"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+)
+
+// TestKarpenterMetricsExist verifies that each expected Karpenter metric is present.
+func TestKarpenterMetricsExist(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			ctx := context.Background()
+			results, err := queryCache.Get(ctx, metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available (is Karpenter installed?)", metricName)
+		})
+	}
+}
+
+// TestKarpenterInstrumentation verifies instrumentation source for all Karpenter metrics.
+func TestKarpenterInstrumentation(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			ctx := context.Background()
+			results, err := queryCache.Get(ctx, metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				r := r
+				name, ok := r.Labels.Instrumentation["@name"]
+				require.True(t, ok, "%s missing @instrumentation.@name", metricName)
+				require.Equal(t, scopeKarpenter, name, "%s instrumentation name", metricName)
+			}
+		})
+	}
+}
+
+// TestKarpenterClusterName verifies the cluster name resource attribute is set correctly.
+func TestKarpenterClusterName(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			ctx := context.Background()
+			results, err := queryCache.Get(ctx, metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				r := r
+				cluster, ok := r.Labels.Resource["k8s.cluster.name"]
+				require.True(t, ok, "%s missing @resource.k8s.cluster.name", metricName)
+				require.Equal(t, cfg.ClusterName, cluster, "%s cluster name mismatch", metricName)
+			}
+		})
+	}
+}
+
+// TestKarpenterExpectedLabels verifies that expected datapoint labels are present.
+func TestKarpenterExpectedLabels(t *testing.T) {
+	t.Parallel()
+	for _, md := range karpenterMetrics {
+		md := md
+		if len(md.ExpectedLabels) == 0 {
+			continue
+		}
+		t.Run(md.Name, func(t *testing.T) {
+			t.Parallel()
+			ctx := context.Background()
+			results, err := queryCache.Get(ctx, md.Name)
+			require.NoError(t, err, "querying %s", md.Name)
+			require.NotEmpty(t, results, "%s not available", md.Name)
+			for _, r := range results {
+				r := r
+				for _, label := range md.ExpectedLabels {
+					label := label
+					_, ok := r.Labels.Datapoint[label]
+					require.True(t, ok, "%s missing expected label '%s'", md.Name, label)
+				}
+			}
+		})
+	}
+}
+
+// TestKarpenterNamespace verifies that Karpenter metrics have a namespace resource attribute.
+func TestKarpenterNamespace(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			ctx := context.Background()
+			results, err := queryCache.Get(ctx, metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				r := r
+				ns, ok := r.Labels.Resource["k8s.namespace.name"]
+				require.True(t, ok, "%s missing @resource.k8s.namespace.name", metricName)
+				require.NotEmpty(t, ns, "%s empty k8s.namespace.name", metricName)
+			}
+		})
+	}
+}
diff --git a/test/otel/integrations/karpenter/metrics_test.go b/test/otel/integrations/karpenter/metrics_test.go
new file mode 100644
index 000000000..f332f8f74
--- /dev/null
+++ b/test/otel/integrations/karpenter/metrics_test.go
@@ -0,0 +1,63 @@
+//go:build integration
+
+// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
+// SPDX-License-Identifier: MIT
+
+package karpenter
+
+import "github.com/aws/amazon-cloudwatch-agent-test/util/otelmetrics"
+
+// karpenterMetrics defines the Karpenter controller metrics expected in CloudWatch.
+// These are Prometheus metrics scraped from the Karpenter controller pods and
+// forwarded by the CloudWatch Agent's Karpenter integration pipeline.
+var karpenterMetrics = []otelmetrics.MetricDefinition{
+	// Build info
+	{Name: "karpenter_build_info", MetricType: "gauge", Scope: otelmetrics.ScopeCluster},
+
+	// Pod scheduling metrics
+	{Name: "karpenter_pods_state", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"phase"}},
+	{Name: "karpenter_pods_bound_duration_seconds", MetricType: "histogram", Scope: otelmetrics.ScopeCluster, Unit: "s"},
+
+	// Scheduler metrics
+	{Name: "karpenter_scheduler_ignored_pods_count", MetricType: "gauge", Scope: otelmetrics.ScopeCluster},
+	{Name: "karpenter_scheduler_scheduling_duration_seconds", MetricType: "histogram", Scope: otelmetrics.ScopeCluster, Unit: "s"},
+
+	// Node resource metrics
+	{Name: "karpenter_nodes_allocatable", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},
+	{Name: "karpenter_nodes_total_daemon_requests", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},
+	{Name: "karpenter_nodes_total_daemon_limits", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},
+	{Name: "karpenter_nodes_total_pod_requests", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},
+	{Name: "karpenter_nodes_total_pod_limits", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},
+	{Name: "karpenter_nodes_system_overhead", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},
+	{Name: "karpenter_nodes_current_lifetime_seconds", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, Unit: "s"},
+
+	// Cluster state metrics
+	{Name: "karpenter_cluster_state_node_count", MetricType: "gauge", Scope: otelmetrics.ScopeCluster},
+	{Name: "karpenter_cluster_state_synced", MetricType: "gauge", Scope: otelmetrics.ScopeCluster},
+	{Name: "karpenter_cluster_state_unsynced_time_seconds", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, Unit: "s"},
+	{Name: "karpenter_cluster_utilization_percent", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},
+
+	// NodePool metrics
+	{Name: "karpenter_nodepools_limit", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},
+	{Name: "karpenter_nodepools_usage", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},
+
+	// Cloudprovider metrics
+	{Name: "karpenter_cloudprovider_duration_seconds", MetricType: "histogram", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"method"}, Unit: "s"},
+	{Name: "karpenter_cloudprovider_instance_type_cpu_cores", MetricType: "gauge", Scope: otelmetrics.ScopeCluster},
+	{Name: "karpenter_cloudprovider_instance_type_memory_bytes", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, Unit: "By"},
+	{Name: "karpenter_cloudprovider_instance_type_offering_available", MetricType: "gauge", Scope: otelmetrics.ScopeCluster},
+	{Name: "karpenter_cloudprovider_instance_type_offering_price_estimate", MetricType: "gauge", Scope: otelmetrics.ScopeCluster},
+
+	// Disruption metrics
+	{Name: "karpenter_voluntary_disruption_eligible_nodes", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"reason"}},
+	{Name: "karpenter_voluntary_disruption_consolidation_timeouts_total", MetricType: "counter", Scope: otelmetrics.ScopeCluster, Unit: "1"},
+	{Name: "karpenter_voluntary_disruption_decision_evaluation_duration_seconds", MetricType: "histogram", Scope: otelmetrics.ScopeCluster, Unit: "s"},
+}
+
+func karpenterMetricNames() []string {
+	names := make([]string, len(karpenterMetrics))
+	for i, d := range karpenterMetrics {
+		names[i] = d.Name
+	}
+	return names
+}
diff --git a/test/otel/integrations/karpenter/setup_test.go b/test/otel/integrations/karpenter/setup_test.go
new file mode 100644
index 000000000..c3f437853
--- /dev/null
+++ b/test/otel/integrations/karpenter/setup_test.go
@@ -0,0 +1,101 @@
+//go:build integration
+
+// Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved.
+// SPDX-License-Identifier: MIT
+
+package karpenter
+
+import (
+	"context"
+	"flag"
+	"fmt"
+	"os"
+	"testing"
+	"time"
+
+	awsconfig "github.com/aws/aws-sdk-go-v2/config"
+	"github.com/aws/aws-sdk-go-v2/service/sts"
+
+	"github.com/aws/amazon-cloudwatch-agent-test/environment"
+	"github.com/aws/amazon-cloudwatch-agent-test/util/otelmetrics"
+)
+
+var (
+	cfg        otelmetrics.TestConfig
+	client     *otelmetrics.OtelMetricsClient
+	queryCache *otelmetrics.QueryCache
+)
+
+// Instrumentation scope name constant.
+const scopeKarpenter = "github.com/aws/karpenter"
+
+// Instance types in the Karpenter cluster.
+var clusterHostTypes = []string{"t3.medium"}
+
+func TestMain(m *testing.M) {
+	environment.RegisterEnvironmentMetaDataFlags()
+	flag.Parse()
+	env := environment.GetEnvironmentMetaData()
+
+	region := env.Region
+	if region == "" {
+		region = os.Getenv("AWS_REGION")
+	}
+	if region == "" {
+		fmt.Fprintf(os.Stderr, "Region not set\n")
+		os.Exit(1)
+	}
+
+	clusterName := env.EKSClusterName
+	if clusterName == "" {
+		clusterName = os.Getenv("CLUSTER_NAME")
+	}
+	if clusterName == "" {
+		fmt.Fprintf(os.Stderr, "Cluster name not set\n")
+		os.Exit(1)
+	}
+
+	ctx := context.Background()
+	awsCfg, err := awsconfig.LoadDefaultConfig(ctx, awsconfig.WithRegion(region))
+	if err != nil {
+		fmt.Fprintf(os.Stderr, "AWS config error: %v\n", err)
+		os.Exit(1)
+	}
+	stsClient := sts.NewFromConfig(awsCfg)
+	identity, err := stsClient.GetCallerIdentity(ctx, &sts.GetCallerIdentityInput{})
+	if err != nil {
+		fmt.Fprintf(os.Stderr, "STS GetCallerIdentity error: %v\n", err)
+		os.Exit(1)
+	}
+
+	cfg = otelmetrics.TestConfig{
+		Region:         region,
+		Endpoint:       fmt.Sprintf("https://monitoring.%s.amazonaws.com", region),
+		Timeout:        30 * time.Second,
+		MaxRetries:     3,
+		ClusterName:    clusterName,
+		AccountID:      *identity.Account,
+		SigningService: "monitoring",
+	}
+
+	client, err = otelmetrics.NewClient(ctx, cfg)
+	if err != nil {
+		fmt.Fprintf(os.Stderr, "Client error: %v\n", err)
+		os.Exit(1)
+	}
+
+	hostMappings := []otelmetrics.SourceHostMapping{
+		{Source: otelmetrics.SourceKarpenter, HostTypes: nil},
+	}
+
+	registry := otelmetrics.NewSourceRegistry(clusterHostTypes, hostMappings,
+		otelmetrics.SourceMapping{Source: otelmetrics.SourceKarpenter, Metrics: karpenterMetrics},
+	)
+
+	queryCache = otelmetrics.NewQueryCache(client, cfg.ClusterName,
+		otelmetrics.WithHostTypes(clusterHostTypes),
+		otelmetrics.WithSourceRegistry(registry),
+	)
+
+	os.Exit(m.Run())
+}
diff --git a/util/otelmetrics/source_registry.go b/util/otelmetrics/source_registry.go
index 4e91a4177..075793c37 100644
--- a/util/otelmetrics/source_registry.go
+++ b/util/otelmetrics/source_registry.go
@@ -18,6 +18,7 @@ const (
 	SourceControlPlane
 	SourceKubeStateMetrics
 	SourceKSMNodeScoped
+	SourceKarpenter
 )
 
 // SourceMapping pairs a MetricSource with its metric definitions.

From eb630c05f4e5b584b661c909557d93cbd2567f0e Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Thu, 25 Jun 2026 13:38:40 +0100
Subject: [PATCH 02/14] revert locals refactor, keep original cluster name
 inline

---
 terraform/eks/daemon/otel/main.tf | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/terraform/eks/daemon/otel/main.tf b/terraform/eks/daemon/otel/main.tf
index 5a1097cd6..e2132779a 100644
--- a/terraform/eks/daemon/otel/main.tf
+++ b/terraform/eks/daemon/otel/main.tf
@@ -13,14 +13,13 @@ module "basic_components" {
 }
 
 locals {
-  aws_eks      = "aws eks --region ${var.region}"
-  cluster_name = "cwagent-eks-integ-${module.common.testing_id}"
+  aws_eks = "aws eks --region ${var.region}"
 }
 
 data "aws_partition" "current" {}
 
 resource "aws_eks_cluster" "this" {
-  name     = local.cluster_name
+  name     = "cwagent-eks-integ-${module.common.testing_id}"
   role_arn = module.basic_components.role_arn
   version  = var.k8s_version
   vpc_config {

From 7d1c3b64ab75b9471fc6092d164420f1154e26bd Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Thu, 25 Jun 2026 14:28:31 +0100
Subject: [PATCH 03/14] removed data "aws_partition" "current" from main.tf

---
 terraform/eks/daemon/otel/main.tf | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/terraform/eks/daemon/otel/main.tf b/terraform/eks/daemon/otel/main.tf
index e2132779a..7849d2599 100644
--- a/terraform/eks/daemon/otel/main.tf
+++ b/terraform/eks/daemon/otel/main.tf
@@ -16,8 +16,6 @@ locals {
   aws_eks = "aws eks --region ${var.region}"
 }
 
-data "aws_partition" "current" {}
-
 resource "aws_eks_cluster" "this" {
   name     = "cwagent-eks-integ-${module.common.testing_id}"
   role_arn = module.basic_components.role_arn
@@ -171,7 +169,7 @@ resource "aws_iam_role_policy" "karpenter_controller" {
         Sid      = "SSMGetParameter"
         Effect   = "Allow"
         Action   = "ssm:GetParameter"
-        Resource = "arn:${data.aws_partition.current.partition}:ssm:${var.region}::parameter/aws/service/*"
+        Resource = "arn:aws:ssm:${var.region}::parameter/aws/service/*"
       },
       {
         Sid      = "PricingPermissions"

From 28cf6b5ebbf364c9224c7a54f6895d7a1b25f0a6 Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Thu, 25 Jun 2026 17:25:32 +0100
Subject: [PATCH 04/14] test: added instrumentation consistency check

---
 .../integrations/karpenter/karpenter_test.go  | 23 +++++++++++++++++++
 1 file changed, 23 insertions(+)

diff --git a/test/otel/integrations/karpenter/karpenter_test.go b/test/otel/integrations/karpenter/karpenter_test.go
index b3e86b52e..0483c34f5 100644
--- a/test/otel/integrations/karpenter/karpenter_test.go
+++ b/test/otel/integrations/karpenter/karpenter_test.go
@@ -115,3 +115,26 @@ func TestKarpenterNamespace(t *testing.T) {
 		})
 	}
 }
+
+// TestKarpenterInstrumentationConsistent verifies all data points for a metric
+// report the same instrumentation scope (no mixed sources).
+func TestKarpenterInstrumentationConsistent(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			ctx := context.Background()
+			results, err := queryCache.Get(ctx, metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			names := make(map[string]struct{})
+			for _, r := range results {
+				if n, ok := r.Labels.Instrumentation["@name"]; ok {
+					names[n] = struct{}{}
+				}
+			}
+			require.Equal(t, 1, len(names), "%s has %d distinct instrumentation names", metricName, len(names))
+		})
+	}
+}

From cf9a4f4acd02b896741740d6ca8424b20f9d536b Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Fri, 26 Jun 2026 13:31:33 +0100
Subject: [PATCH 05/14] edited comment

---
 terraform/eks/daemon/otel/main.tf | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/terraform/eks/daemon/otel/main.tf b/terraform/eks/daemon/otel/main.tf
index 7849d2599..54c48b7aa 100644
--- a/terraform/eks/daemon/otel/main.tf
+++ b/terraform/eks/daemon/otel/main.tf
@@ -110,9 +110,7 @@ resource "aws_eks_addon" "pod_identity_agent" {
   addon_name   = "eks-pod-identity-agent"
 }
 
-# -----------------------------------------------------------------------------
-# Karpenter Controller IAM Role (Pod Identity)
-# -----------------------------------------------------------------------------
+# --- Karpenter Controller IAM Role (Pod Identity) ---
 
 resource "aws_iam_role" "karpenter_controller" {
   name = "cwagent-otel-karpenter-${module.common.testing_id}"

From ec8908712c9699508492530c29929f05ec954b32 Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Fri, 26 Jun 2026 16:21:17 +0100
Subject: [PATCH 06/14] refactor: consolidate resource attribute tests into
 single test

---
 .../integrations/karpenter/karpenter_test.go  | 60 +++++++++----------
 1 file changed, 27 insertions(+), 33 deletions(-)

diff --git a/test/otel/integrations/karpenter/karpenter_test.go b/test/otel/integrations/karpenter/karpenter_test.go
index 0483c34f5..ceecc8f15 100644
--- a/test/otel/integrations/karpenter/karpenter_test.go
+++ b/test/otel/integrations/karpenter/karpenter_test.go
@@ -48,8 +48,9 @@ func TestKarpenterInstrumentation(t *testing.T) {
 	}
 }
 
-// TestKarpenterClusterName verifies the cluster name resource attribute is set correctly.
-func TestKarpenterClusterName(t *testing.T) {
+// TestKarpenterInstrumentationConsistent verifies all data points for a metric
+// report the same instrumentation scope (no mixed sources).
+func TestKarpenterInstrumentationConsistent(t *testing.T) {
 	t.Parallel()
 	for _, metricName := range karpenterMetricNames() {
 		metricName := metricName
@@ -59,12 +60,13 @@ func TestKarpenterClusterName(t *testing.T) {
 			results, err := queryCache.Get(ctx, metricName)
 			require.NoError(t, err, "querying %s", metricName)
 			require.NotEmpty(t, results, "%s not available", metricName)
+			names := make(map[string]struct{})
 			for _, r := range results {
-				r := r
-				cluster, ok := r.Labels.Resource["k8s.cluster.name"]
-				require.True(t, ok, "%s missing @resource.k8s.cluster.name", metricName)
-				require.Equal(t, cfg.ClusterName, cluster, "%s cluster name mismatch", metricName)
+				if n, ok := r.Labels.Instrumentation["@name"]; ok {
+					names[n] = struct{}{}
+				}
 			}
+			require.Equal(t, 1, len(names), "%s has %d distinct instrumentation names", metricName, len(names))
 		})
 	}
 }
@@ -95,31 +97,19 @@ func TestKarpenterExpectedLabels(t *testing.T) {
 	}
 }
 
-// TestKarpenterNamespace verifies that Karpenter metrics have a namespace resource attribute.
-func TestKarpenterNamespace(t *testing.T) {
+// TestKarpenterResourceAttributes verifies K8s and cloud resource attributes are enriched
+// correctly, including cluster name value validation.
+func TestKarpenterResourceAttributes(t *testing.T) {
 	t.Parallel()
-	for _, metricName := range karpenterMetricNames() {
-		metricName := metricName
-		t.Run(metricName, func(t *testing.T) {
-			t.Parallel()
-			ctx := context.Background()
-			results, err := queryCache.Get(ctx, metricName)
-			require.NoError(t, err, "querying %s", metricName)
-			require.NotEmpty(t, results, "%s not available", metricName)
-			for _, r := range results {
-				r := r
-				ns, ok := r.Labels.Resource["k8s.namespace.name"]
-				require.True(t, ok, "%s missing @resource.k8s.namespace.name", metricName)
-				require.NotEmpty(t, ns, "%s empty k8s.namespace.name", metricName)
-			}
-		})
+	requiredAttrs := []string{
+		"k8s.pod.name",
+		"k8s.deployment.name",
+		"k8s.namespace.name",
+		"k8s.cluster.name",
+		"cloud.provider",
+		"cloud.region",
+		"cloud.platform",
 	}
-}
-
-// TestKarpenterInstrumentationConsistent verifies all data points for a metric
-// report the same instrumentation scope (no mixed sources).
-func TestKarpenterInstrumentationConsistent(t *testing.T) {
-	t.Parallel()
 	for _, metricName := range karpenterMetricNames() {
 		metricName := metricName
 		t.Run(metricName, func(t *testing.T) {
@@ -128,13 +118,17 @@ func TestKarpenterInstrumentationConsistent(t *testing.T) {
 			results, err := queryCache.Get(ctx, metricName)
 			require.NoError(t, err, "querying %s", metricName)
 			require.NotEmpty(t, results, "%s not available", metricName)
-			names := make(map[string]struct{})
 			for _, r := range results {
-				if n, ok := r.Labels.Instrumentation["@name"]; ok {
-					names[n] = struct{}{}
+				r := r
+				for _, attr := range requiredAttrs {
+					v, ok := r.Labels.Resource[attr]
+					require.True(t, ok, "%s missing @resource.%s", metricName, attr)
+					require.NotEmpty(t, v, "%s empty @resource.%s", metricName, attr)
 				}
+				// Validate cluster name matches the expected value
+				require.Equal(t, cfg.ClusterName, r.Labels.Resource["k8s.cluster.name"],
+					"%s cluster name mismatch", metricName)
 			}
-			require.Equal(t, 1, len(names), "%s has %d distinct instrumentation names", metricName, len(names))
 		})
 	}
 }

From db96765905d7fd9afa243d6ac118d0d48dc24c96 Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Thu, 2 Jul 2026 15:14:09 +0100
Subject: [PATCH 07/14] test(karpenter): Added value assertions for cloud and
 K8s resource attributes

---
 .../integrations/karpenter/karpenter_test.go  | 263 ++++++++++++++++--
 1 file changed, 237 insertions(+), 26 deletions(-)

diff --git a/test/otel/integrations/karpenter/karpenter_test.go b/test/otel/integrations/karpenter/karpenter_test.go
index ceecc8f15..6410ff0a8 100644
--- a/test/otel/integrations/karpenter/karpenter_test.go
+++ b/test/otel/integrations/karpenter/karpenter_test.go
@@ -7,12 +7,22 @@ package karpenter
 
 import (
 	"context"
+	"fmt"
+	"strings"
 	"testing"
 
 	"github.com/stretchr/testify/require"
 )
 
-// TestKarpenterMetricsExist verifies that each expected Karpenter metric is present.
+// ===========================================================================
+// Metric Existence Tests
+// ===========================================================================
+
+// ---------------------------------------------------------------------------
+// TestKarpenterMetricsExist — verify each expected metric is present in
+// CloudWatch.
+// ---------------------------------------------------------------------------
+
 func TestKarpenterMetricsExist(t *testing.T) {
 	t.Parallel()
 	for _, metricName := range karpenterMetricNames() {
@@ -27,7 +37,15 @@ func TestKarpenterMetricsExist(t *testing.T) {
 	}
 }
 
-// TestKarpenterInstrumentation verifies instrumentation source for all Karpenter metrics.
+// ===========================================================================
+// Instrumentation Scope Tests
+// ===========================================================================
+
+// ---------------------------------------------------------------------------
+// TestKarpenterInstrumentation — verify instrumentation scope name for all
+// Karpenter metrics.
+// ---------------------------------------------------------------------------
+
 func TestKarpenterInstrumentation(t *testing.T) {
 	t.Parallel()
 	for _, metricName := range karpenterMetricNames() {
@@ -48,8 +66,11 @@ func TestKarpenterInstrumentation(t *testing.T) {
 	}
 }
 
-// TestKarpenterInstrumentationConsistent verifies all data points for a metric
-// report the same instrumentation scope (no mixed sources).
+// ---------------------------------------------------------------------------
+// TestKarpenterInstrumentationConsistent — verify all data points for a
+// metric report the same instrumentation scope (no mixed sources).
+// ---------------------------------------------------------------------------
+
 func TestKarpenterInstrumentationConsistent(t *testing.T) {
 	t.Parallel()
 	for _, metricName := range karpenterMetricNames() {
@@ -71,7 +92,15 @@ func TestKarpenterInstrumentationConsistent(t *testing.T) {
 	}
 }
 
-// TestKarpenterExpectedLabels verifies that expected datapoint labels are present.
+// ===========================================================================
+// Datapoint Label Tests
+// ===========================================================================
+
+// ---------------------------------------------------------------------------
+// TestKarpenterExpectedLabels — verify expected datapoint labels are present
+// on metrics that declare them.
+// ---------------------------------------------------------------------------
+
 func TestKarpenterExpectedLabels(t *testing.T) {
 	t.Parallel()
 	for _, md := range karpenterMetrics {
@@ -97,37 +126,219 @@ func TestKarpenterExpectedLabels(t *testing.T) {
 	}
 }
 
-// TestKarpenterResourceAttributes verifies K8s and cloud resource attributes are enriched
-// correctly, including cluster name value validation.
-func TestKarpenterResourceAttributes(t *testing.T) {
+// ===========================================================================
+// K8s Resource Attribute Tests
+// ===========================================================================
+
+// ---------------------------------------------------------------------------
+// TestKarpenterClusterIdentity — k8s.cluster.name must match the configured
+// cluster.
+// ---------------------------------------------------------------------------
+
+func TestKarpenterClusterIdentity(t *testing.T) {
 	t.Parallel()
-	requiredAttrs := []string{
-		"k8s.pod.name",
-		"k8s.deployment.name",
-		"k8s.namespace.name",
-		"k8s.cluster.name",
-		"cloud.provider",
-		"cloud.region",
-		"cloud.platform",
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			results, err := queryCache.Get(context.Background(), metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				clusterName, ok := r.Labels.Resource["k8s.cluster.name"]
+				require.True(t, ok, "%s missing @resource.k8s.cluster.name", metricName)
+				require.Equal(t, cfg.ClusterName, clusterName, "%s k8s.cluster.name", metricName)
+			}
+		})
 	}
+}
+
+// ---------------------------------------------------------------------------
+// TestKarpenterNamespace — all Karpenter metrics must originate from the
+// kube-system namespace.
+// ---------------------------------------------------------------------------
+
+func TestKarpenterNamespace(t *testing.T) {
+	t.Parallel()
 	for _, metricName := range karpenterMetricNames() {
 		metricName := metricName
 		t.Run(metricName, func(t *testing.T) {
 			t.Parallel()
-			ctx := context.Background()
-			results, err := queryCache.Get(ctx, metricName)
+			results, err := queryCache.Get(context.Background(), metricName)
 			require.NoError(t, err, "querying %s", metricName)
 			require.NotEmpty(t, results, "%s not available", metricName)
 			for _, r := range results {
-				r := r
-				for _, attr := range requiredAttrs {
-					v, ok := r.Labels.Resource[attr]
-					require.True(t, ok, "%s missing @resource.%s", metricName, attr)
-					require.NotEmpty(t, v, "%s empty @resource.%s", metricName, attr)
+				ns, ok := r.Labels.Resource["k8s.namespace.name"]
+				require.True(t, ok, "%s missing @resource.k8s.namespace.name", metricName)
+				require.Equal(t, "kube-system", ns, "%s k8s.namespace.name", metricName)
+			}
+		})
+	}
+}
+
+// ---------------------------------------------------------------------------
+// TestKarpenterDeploymentName — k8s.deployment.name must be "karpenter".
+// ---------------------------------------------------------------------------
+
+func TestKarpenterDeploymentName(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			results, err := queryCache.Get(context.Background(), metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				deploy, ok := r.Labels.Resource["k8s.deployment.name"]
+				require.True(t, ok, "%s missing @resource.k8s.deployment.name", metricName)
+				require.Equal(t, "karpenter", deploy, "%s k8s.deployment.name", metricName)
+			}
+		})
+	}
+}
+
+// ---------------------------------------------------------------------------
+// TestKarpenterPodName — k8s.pod.name must start with "karpenter-".
+// ---------------------------------------------------------------------------
+
+func TestKarpenterPodName(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			results, err := queryCache.Get(context.Background(), metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				podName, ok := r.Labels.Resource["k8s.pod.name"]
+				require.True(t, ok, "%s missing @resource.k8s.pod.name", metricName)
+				require.True(t, strings.HasPrefix(podName, "karpenter-"),
+					"%s k8s.pod.name should start with 'karpenter-', got %q", metricName, podName)
+			}
+		})
+	}
+}
+
+// ===========================================================================
+// Cloud Resource Attribute Tests
+// ===========================================================================
+
+// ---------------------------------------------------------------------------
+// TestKarpenterCloudProvider — cloud.provider must be "aws".
+// ---------------------------------------------------------------------------
+
+func TestKarpenterCloudProvider(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			results, err := queryCache.Get(context.Background(), metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				provider, ok := r.Labels.Resource["cloud.provider"]
+				require.True(t, ok, "%s missing @resource.cloud.provider", metricName)
+				require.Equal(t, "aws", provider, "%s cloud.provider", metricName)
+			}
+		})
+	}
+}
+
+// ---------------------------------------------------------------------------
+// TestKarpenterCloudPlatform — cloud.platform must be "aws_eks".
+// ---------------------------------------------------------------------------
+
+func TestKarpenterCloudPlatform(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			results, err := queryCache.Get(context.Background(), metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				platform, ok := r.Labels.Resource["cloud.platform"]
+				require.True(t, ok, "%s missing @resource.cloud.platform", metricName)
+				require.Equal(t, "aws_eks", platform, "%s cloud.platform", metricName)
+			}
+		})
+	}
+}
+
+// ---------------------------------------------------------------------------
+// TestKarpenterCloudRegion — cloud.region must match the configured region.
+// ---------------------------------------------------------------------------
+
+func TestKarpenterCloudRegion(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			results, err := queryCache.Get(context.Background(), metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				region, ok := r.Labels.Resource["cloud.region"]
+				require.True(t, ok, "%s missing @resource.cloud.region", metricName)
+				require.Equal(t, cfg.Region, region, "%s cloud.region", metricName)
+			}
+		})
+	}
+}
+
+// ---------------------------------------------------------------------------
+// TestKarpenterCloudAccountID — cloud.account.id must match the test account.
+// ---------------------------------------------------------------------------
+
+func TestKarpenterCloudAccountID(t *testing.T) {
+	t.Parallel()
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			results, err := queryCache.Get(context.Background(), metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				acctID, ok := r.Labels.Resource["cloud.account.id"]
+				if ok {
+					require.Equal(t, cfg.AccountID, acctID, "%s cloud.account.id", metricName)
+				}
+			}
+		})
+	}
+}
+
+// ---------------------------------------------------------------------------
+// TestKarpenterCloudResourceID — cloud.resource_id must be a valid EKS
+// cluster ARN.
+// ---------------------------------------------------------------------------
+
+func TestKarpenterCloudResourceID(t *testing.T) {
+	t.Parallel()
+	expectedARNPrefix := fmt.Sprintf("arn:aws:eks:%s:", cfg.Region)
+	expectedARNSuffix := fmt.Sprintf(":cluster/%s", cfg.ClusterName)
+
+	for _, metricName := range karpenterMetricNames() {
+		metricName := metricName
+		t.Run(metricName, func(t *testing.T) {
+			t.Parallel()
+			results, err := queryCache.Get(context.Background(), metricName)
+			require.NoError(t, err, "querying %s", metricName)
+			require.NotEmpty(t, results, "%s not available", metricName)
+			for _, r := range results {
+				arn, ok := r.Labels.Resource["cloud.resource_id"]
+				if ok {
+					require.True(t, strings.HasPrefix(arn, expectedARNPrefix),
+						"%s cloud.resource_id should start with %q, got %q", metricName, expectedARNPrefix, arn)
+					require.True(t, strings.HasSuffix(arn, expectedARNSuffix),
+						"%s cloud.resource_id should end with %q, got %q", metricName, expectedARNSuffix, arn)
 				}
-				// Validate cluster name matches the expected value
-				require.Equal(t, cfg.ClusterName, r.Labels.Resource["k8s.cluster.name"],
-					"%s cluster name mismatch", metricName)
 			}
 		})
 	}

From a1009f483bc6f504ebb8b323eff7362fb18452a4 Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Mon, 6 Jul 2026 14:52:54 +0100
Subject: [PATCH 08/14] fix : renamed flag name in main.tf

---
 terraform/eks/daemon/otel/main.tf | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/terraform/eks/daemon/otel/main.tf b/terraform/eks/daemon/otel/main.tf
index 54c48b7aa..c4dad1401 100644
--- a/terraform/eks/daemon/otel/main.tf
+++ b/terraform/eks/daemon/otel/main.tf
@@ -344,7 +344,7 @@ resource "helm_release" "aws_observability" {
     { name = "clusterName", value = aws_eks_cluster.this.name },
     { name = "region", value = var.region },
     { name = "otelContainerInsights.enabled", value = "true" },
-    { name = "otelContainerInsights.metrics.karpenter.enabled", value = "true" },
+    { name = "otelContainerInsights.integrations.karpenter.enabled", value = "true" },
   ]
 
   depends_on = [

From 15652d2bbbfdda85b0645357fea43f8cf444647e Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Wed, 8 Jul 2026 13:12:12 +0100
Subject: [PATCH 09/14] fix: update karpenter to 1.13.0, remove SQS queue

---
 terraform/eks/daemon/otel/main.tf      | 31 --------------------------
 terraform/eks/daemon/otel/variables.tf |  2 +-
 2 files changed, 1 insertion(+), 32 deletions(-)

diff --git a/terraform/eks/daemon/otel/main.tf b/terraform/eks/daemon/otel/main.tf
index c4dad1401..e49ecb518 100644
--- a/terraform/eks/daemon/otel/main.tf
+++ b/terraform/eks/daemon/otel/main.tf
@@ -175,40 +175,10 @@ resource "aws_iam_role_policy" "karpenter_controller" {
         Action   = "pricing:GetProducts"
         Resource = "*"
       },
-      {
-        Sid    = "SQSPermissions"
-        Effect = "Allow"
-        Action = [
-          "sqs:DeleteMessage",
-          "sqs:GetQueueUrl",
-          "sqs:ReceiveMessage",
-        ]
-        Resource = aws_sqs_queue.karpenter_interruption.arn
-      },
     ]
   })
 }
 
-# Karpenter Interruption Queue (SQS)
-resource "aws_sqs_queue" "karpenter_interruption" {
-  name                      = "cwagent-otel-karpenter-${module.common.testing_id}"
-  message_retention_seconds = 300
-  sqs_managed_sse_enabled   = true
-}
-
-resource "aws_sqs_queue_policy" "karpenter_interruption" {
-  queue_url = aws_sqs_queue.karpenter_interruption.id
-  policy = jsonencode({
-    Version = "2012-10-17"
-    Statement = [{
-      Effect    = "Allow"
-      Principal = { Service = ["events.amazonaws.com", "sqs.amazonaws.com"] }
-      Action    = "sqs:SendMessage"
-      Resource  = aws_sqs_queue.karpenter_interruption.arn
-    }]
-  })
-}
-
 # Karpenter Pod Identity Association
 resource "aws_eks_pod_identity_association" "karpenter_controller" {
   depends_on      = [aws_eks_addon.pod_identity_agent]
@@ -254,7 +224,6 @@ resource "helm_release" "karpenter" {
   set = [
     { name = "settings.clusterName", value = aws_eks_cluster.this.name },
     { name = "settings.clusterEndpoint", value = aws_eks_cluster.this.endpoint },
-    { name = "settings.interruptionQueue", value = aws_sqs_queue.karpenter_interruption.name },
     { name = "controller.resources.requests.cpu", value = "100m" },
     { name = "controller.resources.requests.memory", value = "256Mi" },
   ]
diff --git a/terraform/eks/daemon/otel/variables.tf b/terraform/eks/daemon/otel/variables.tf
index 7ed580859..095e2e6ce 100644
--- a/terraform/eks/daemon/otel/variables.tf
+++ b/terraform/eks/daemon/otel/variables.tf
@@ -43,5 +43,5 @@ variable "instance_type" {
 
 variable "karpenter_version" {
   type    = string
-  default = "1.4.0"
+  default = "1.13.0"
 }

From 260dda526bb2fb9e8d860d7433b3ba3064d01f6f Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Wed, 8 Jul 2026 14:31:47 +0100
Subject: [PATCH 10/14] ci: temporarily point helm chart to fork for testing

---
 terraform/eks/daemon/otel/main.tf      | 2 +-
 terraform/eks/daemon/otel/variables.tf | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/terraform/eks/daemon/otel/main.tf b/terraform/eks/daemon/otel/main.tf
index e49ecb518..9b5283603 100644
--- a/terraform/eks/daemon/otel/main.tf
+++ b/terraform/eks/daemon/otel/main.tf
@@ -297,7 +297,7 @@ resource "null_resource" "karpenter_nodepool" {
 data "external" "clone_helm_chart" {
   program = ["bash", "-c", <<-EOT
     rm -rf ./helm-charts
-    git clone -b ${var.helm_chart_branch} https://github.com/aws-observability/helm-charts.git ./helm-charts
+    git clone -b ${var.helm_chart_branch} https://github.com/vaishnavi-30-beep/helm-charts.git ./helm-charts
     echo '{"status":"ready"}'
   EOT
   ]
diff --git a/terraform/eks/daemon/otel/variables.tf b/terraform/eks/daemon/otel/variables.tf
index 095e2e6ce..33b56a0af 100644
--- a/terraform/eks/daemon/otel/variables.tf
+++ b/terraform/eks/daemon/otel/variables.tf
@@ -23,7 +23,7 @@ variable "cwagent_image_tag" {
 
 variable "helm_chart_branch" {
   type    = string
-  default = "main"
+  default = "feature/karpenter-integration"
 }
 
 variable "k8s_version" {

From 556c4fc2f0dc5d5bd22bfdf8a40c8bb663d8ea13 Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Wed, 8 Jul 2026 19:01:19 +0100
Subject: [PATCH 11/14] fix: remove scheduler metrics that require pod
 scheduling activity

---
 test/otel/integrations/karpenter/metrics_test.go | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/test/otel/integrations/karpenter/metrics_test.go b/test/otel/integrations/karpenter/metrics_test.go
index f332f8f74..77639fbce 100644
--- a/test/otel/integrations/karpenter/metrics_test.go
+++ b/test/otel/integrations/karpenter/metrics_test.go
@@ -18,10 +18,6 @@ var karpenterMetrics = []otelmetrics.MetricDefinition{
 	{Name: "karpenter_pods_state", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"phase"}},
 	{Name: "karpenter_pods_bound_duration_seconds", MetricType: "histogram", Scope: otelmetrics.ScopeCluster, Unit: "s"},
 
-	// Scheduler metrics
-	{Name: "karpenter_scheduler_ignored_pods_count", MetricType: "gauge", Scope: otelmetrics.ScopeCluster},
-	{Name: "karpenter_scheduler_scheduling_duration_seconds", MetricType: "histogram", Scope: otelmetrics.ScopeCluster, Unit: "s"},
-
 	// Node resource metrics
 	{Name: "karpenter_nodes_allocatable", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},
 	{Name: "karpenter_nodes_total_daemon_requests", MetricType: "gauge", Scope: otelmetrics.ScopeCluster, ExpectedLabels: []string{"resource_type"}},

From 2270369ba1e5a07571fe6cde2b1318a110c8ed0c Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Wed, 8 Jul 2026 19:16:11 +0100
Subject: [PATCH 12/14] fix: rename helm_chart_branch to avoid CI override,
 remove scheduler metrics

---
 terraform/eks/daemon/otel/main.tf      | 2 +-
 terraform/eks/daemon/otel/variables.tf | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/terraform/eks/daemon/otel/main.tf b/terraform/eks/daemon/otel/main.tf
index 9b5283603..90c532215 100644
--- a/terraform/eks/daemon/otel/main.tf
+++ b/terraform/eks/daemon/otel/main.tf
@@ -297,7 +297,7 @@ resource "null_resource" "karpenter_nodepool" {
 data "external" "clone_helm_chart" {
   program = ["bash", "-c", <<-EOT
     rm -rf ./helm-charts
-    git clone -b ${var.helm_chart_branch} https://github.com/vaishnavi-30-beep/helm-charts.git ./helm-charts
+    git clone -b ${var.karpenter_helm_chart_branch} https://github.com/vaishnavi-30-beep/helm-charts.git ./helm-charts
     echo '{"status":"ready"}'
   EOT
   ]
diff --git a/terraform/eks/daemon/otel/variables.tf b/terraform/eks/daemon/otel/variables.tf
index 33b56a0af..a9ed23e6c 100644
--- a/terraform/eks/daemon/otel/variables.tf
+++ b/terraform/eks/daemon/otel/variables.tf
@@ -21,7 +21,7 @@ variable "cwagent_image_tag" {
   default = "latest"
 }
 
-variable "helm_chart_branch" {
+variable "karpenter_helm_chart_branch" {
   type    = string
   default = "feature/karpenter-integration"
 }

From 156d1125a367ca52f349870934e4eff1d953137c Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Thu, 9 Jul 2026 00:08:20 +0100
Subject: [PATCH 13/14] fix: rename to karpenter_chart_ref to avoid CI grep
 override

---
 terraform/eks/daemon/otel/main.tf      | 2 +-
 terraform/eks/daemon/otel/variables.tf | 8 +++++++-
 2 files changed, 8 insertions(+), 2 deletions(-)

diff --git a/terraform/eks/daemon/otel/main.tf b/terraform/eks/daemon/otel/main.tf
index 90c532215..69a2ffa49 100644
--- a/terraform/eks/daemon/otel/main.tf
+++ b/terraform/eks/daemon/otel/main.tf
@@ -297,7 +297,7 @@ resource "null_resource" "karpenter_nodepool" {
 data "external" "clone_helm_chart" {
   program = ["bash", "-c", <<-EOT
     rm -rf ./helm-charts
-    git clone -b ${var.karpenter_helm_chart_branch} https://github.com/vaishnavi-30-beep/helm-charts.git ./helm-charts
+    git clone -b ${var.karpenter_chart_ref} https://github.com/vaishnavi-30-beep/helm-charts.git ./helm-charts
     echo '{"status":"ready"}'
   EOT
   ]
diff --git a/terraform/eks/daemon/otel/variables.tf b/terraform/eks/daemon/otel/variables.tf
index a9ed23e6c..caccb44ba 100644
--- a/terraform/eks/daemon/otel/variables.tf
+++ b/terraform/eks/daemon/otel/variables.tf
@@ -21,11 +21,17 @@ variable "cwagent_image_tag" {
   default = "latest"
 }
 
-variable "karpenter_helm_chart_branch" {
+variable "karpenter_chart_ref" {
   type    = string
   default = "feature/karpenter-integration"
 }
 
+variable "helm_chart_branch" {
+  type    = string
+  default = "main"
+  description = "Passed by CI workflow but unused — karpenter_chart_ref is used instead"
+}
+
 variable "k8s_version" {
   type    = string
   default = "1.35"

From abc9d22985a1c04b47ee4a3dcbca648c4d1b6356 Mon Sep 17 00:00:00 2001
From: Vaishnavi Joshi <vaijo@amazon.com>
Date: Thu, 9 Jul 2026 11:19:23 +0100
Subject: [PATCH 14/14] revert: remove temporary fork/branch overrides

---
 terraform/eks/daemon/otel/main.tf      | 2 +-
 terraform/eks/daemon/otel/variables.tf | 6 ------
 2 files changed, 1 insertion(+), 7 deletions(-)

diff --git a/terraform/eks/daemon/otel/main.tf b/terraform/eks/daemon/otel/main.tf
index 69a2ffa49..e49ecb518 100644
--- a/terraform/eks/daemon/otel/main.tf
+++ b/terraform/eks/daemon/otel/main.tf
@@ -297,7 +297,7 @@ resource "null_resource" "karpenter_nodepool" {
 data "external" "clone_helm_chart" {
   program = ["bash", "-c", <<-EOT
     rm -rf ./helm-charts
-    git clone -b ${var.karpenter_chart_ref} https://github.com/vaishnavi-30-beep/helm-charts.git ./helm-charts
+    git clone -b ${var.helm_chart_branch} https://github.com/aws-observability/helm-charts.git ./helm-charts
     echo '{"status":"ready"}'
   EOT
   ]
diff --git a/terraform/eks/daemon/otel/variables.tf b/terraform/eks/daemon/otel/variables.tf
index caccb44ba..095e2e6ce 100644
--- a/terraform/eks/daemon/otel/variables.tf
+++ b/terraform/eks/daemon/otel/variables.tf
@@ -21,15 +21,9 @@ variable "cwagent_image_tag" {
   default = "latest"
 }
 
-variable "karpenter_chart_ref" {
-  type    = string
-  default = "feature/karpenter-integration"
-}
-
 variable "helm_chart_branch" {
   type    = string
   default = "main"
-  description = "Passed by CI workflow but unused — karpenter_chart_ref is used instead"
 }
 
 variable "k8s_version" {