Merge pull request #14 from cybozu-go/fix-202504

YZ775 · web-flow · commit e5225737b349 · 2025-04-30T13:30:45.000+09:00
fix metrics name and not to list PodMetrics/NodeMetrics
diff --git a/Makefile b/Makefile
@@ -39,6 +39,7 @@ docker-build:
 .PHONY: maintenance
 maintenance: update-tools-versions update-actions
 	$(MAKE) -C ./e2e update-tools-version
+	$(MAKE) -C ./e2e/manifests/metrics-server update-manifests
 
 .PHONY: update-tools-versions
 update-tools-versions: login-gh
diff --git a/cmd/root.go b/cmd/root.go
@@ -3,6 +3,7 @@ package cmd
 import (
 	"context"
 	"errors"
+	"fmt"
 	"log"
 	"os"
 	"time"
@@ -53,6 +54,19 @@ type resourceMetadata struct {
 	deletionTimestamp *metav1.Time
 }
 
+var IgnoreResources = []schema.GroupVersionResource{
+	{
+		Group:    "metrics.k8s.io",
+		Version:  "v1beta1",
+		Resource: "pods",
+	},
+	{
+		Group:    "metrics.k8s.io",
+		Version:  "v1beta1",
+		Resource: "nodes",
+	},
+}
+
 func getAllResources(ctx context.Context, config *rest.Config) ([]resourceMetadata, error) {
 	o, err := discovery.NewDiscoveryClientForConfig(config)
 	if err != nil {
@@ -74,6 +88,12 @@ func getAllResources(ctx context.Context, config *rest.Config) ([]resourceMetada
 		}
 		for _, resource := range resList.APIResources {
 			groupResourceDef := schema.GroupVersionResource{Group: gv.Group, Version: gv.Version, Resource: resource.Name}
+			for _, ir := range IgnoreResources {
+				if ir == groupResourceDef {
+					fmt.Printf("ignoring %s %s %s\n", groupResourceDef.Group, groupResourceDef.Version, groupResourceDef.Resource)
+					continue
+				}
+			}
 			listResponse, err := dynamicClient.Resource(groupResourceDef).Namespace(corev1.NamespaceAll).List(ctx, metav1.ListOptions{})
 			statusErr := &apierrors.StatusError{}
 			if err != nil && !errors.As(err, &statusErr) {
@@ -149,7 +169,7 @@ func postZombieResourcesMetrics(zombieResources []resourceMetadata, endpoint str
 	gauges := make([]prometheus.Gauge, 0)
 	for _, res := range zombieResources {
 		gauge := prometheus.NewGauge(prometheus.GaugeOpts{
-			Name: "zombie_duration_hours",
+			Name: "zombie_duration_seconds",
 			Help: "zombie detector zombie duration",
 			ConstLabels: map[string]string{
 				"apiVersion": res.version,
diff --git a/e2e/Makefile b/e2e/Makefile
@@ -40,6 +40,7 @@ start:
 	$(KUSTOMIZE) build ../config/namespace | $(KUBECTL) apply -f -
 	$(KUSTOMIZE) build ../config/rbac | $(KUBECTL) apply -f -
 	$(KUSTOMIZE) build ./manifests/pushgateway | $(KUBECTL) apply -f -
+	$(KUSTOMIZE) build ./manifests/metrics-server | $(KUBECTL) apply -f -
 
 
 
diff --git a/e2e/manifests/metrics-server/Makefile b/e2e/manifests/metrics-server/Makefile
@@ -0,0 +1,3 @@
+.PHONY: update-manifests
+update-manifests:
+	https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml
diff --git a/e2e/manifests/metrics-server/components.yaml b/e2e/manifests/metrics-server/components.yaml
@@ -0,0 +1,201 @@
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  labels:
+    k8s-app: metrics-server
+  name: metrics-server
+  namespace: kube-system
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRole
+metadata:
+  labels:
+    k8s-app: metrics-server
+    rbac.authorization.k8s.io/aggregate-to-admin: "true"
+    rbac.authorization.k8s.io/aggregate-to-edit: "true"
+    rbac.authorization.k8s.io/aggregate-to-view: "true"
+  name: system:aggregated-metrics-reader
+rules:
+- apiGroups:
+  - metrics.k8s.io
+  resources:
+  - pods
+  - nodes
+  verbs:
+  - get
+  - list
+  - watch
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRole
+metadata:
+  labels:
+    k8s-app: metrics-server
+  name: system:metrics-server
+rules:
+- apiGroups:
+  - ""
+  resources:
+  - nodes/metrics
+  verbs:
+  - get
+- apiGroups:
+  - ""
+  resources:
+  - pods
+  - nodes
+  verbs:
+  - get
+  - list
+  - watch
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: RoleBinding
+metadata:
+  labels:
+    k8s-app: metrics-server
+  name: metrics-server-auth-reader
+  namespace: kube-system
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: Role
+  name: extension-apiserver-authentication-reader
+subjects:
+- kind: ServiceAccount
+  name: metrics-server
+  namespace: kube-system
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRoleBinding
+metadata:
+  labels:
+    k8s-app: metrics-server
+  name: metrics-server:system:auth-delegator
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: ClusterRole
+  name: system:auth-delegator
+subjects:
+- kind: ServiceAccount
+  name: metrics-server
+  namespace: kube-system
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRoleBinding
+metadata:
+  labels:
+    k8s-app: metrics-server
+  name: system:metrics-server
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: ClusterRole
+  name: system:metrics-server
+subjects:
+- kind: ServiceAccount
+  name: metrics-server
+  namespace: kube-system
+---
+apiVersion: v1
+kind: Service
+metadata:
+  labels:
+    k8s-app: metrics-server
+  name: metrics-server
+  namespace: kube-system
+spec:
+  ports:
+  - name: https
+    port: 443
+    protocol: TCP
+    targetPort: https
+  selector:
+    k8s-app: metrics-server
+---
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  labels:
+    k8s-app: metrics-server
+  name: metrics-server
+  namespace: kube-system
+spec:
+  selector:
+    matchLabels:
+      k8s-app: metrics-server
+  strategy:
+    rollingUpdate:
+      maxUnavailable: 0
+  template:
+    metadata:
+      labels:
+        k8s-app: metrics-server
+    spec:
+      containers:
+      - args:
+        - --cert-dir=/tmp
+        - --secure-port=10250
+        - --kubelet-preferred-address-types=InternalIP,ExternalIP,Hostname
+        - --kubelet-use-node-status-port
+        - --metric-resolution=15s
+        image: registry.k8s.io/metrics-server/metrics-server:v0.7.2
+        imagePullPolicy: IfNotPresent
+        livenessProbe:
+          failureThreshold: 3
+          httpGet:
+            path: /livez
+            port: https
+            scheme: HTTPS
+          periodSeconds: 10
+        name: metrics-server
+        ports:
+        - containerPort: 10250
+          name: https
+          protocol: TCP
+        readinessProbe:
+          failureThreshold: 3
+          httpGet:
+            path: /readyz
+            port: https
+            scheme: HTTPS
+          initialDelaySeconds: 20
+          periodSeconds: 10
+        resources:
+          requests:
+            cpu: 100m
+            memory: 200Mi
+        securityContext:
+          allowPrivilegeEscalation: false
+          capabilities:
+            drop:
+            - ALL
+          readOnlyRootFilesystem: true
+          runAsNonRoot: true
+          runAsUser: 1000
+          seccompProfile:
+            type: RuntimeDefault
+        volumeMounts:
+        - mountPath: /tmp
+          name: tmp-dir
+      nodeSelector:
+        kubernetes.io/os: linux
+      priorityClassName: system-cluster-critical
+      serviceAccountName: metrics-server
+      volumes:
+      - emptyDir: {}
+        name: tmp-dir
+---
+apiVersion: apiregistration.k8s.io/v1
+kind: APIService
+metadata:
+  labels:
+    k8s-app: metrics-server
+  name: v1beta1.metrics.k8s.io
+spec:
+  group: metrics.k8s.io
+  groupPriorityMinimum: 100
+  insecureSkipTLSVerify: true
+  service:
+    name: metrics-server
+    namespace: kube-system
+  version: v1beta1
+  versionPriority: 100
diff --git a/e2e/manifests/metrics-server/kustomization.yaml b/e2e/manifests/metrics-server/kustomization.yaml
@@ -0,0 +1,4 @@
+resources:
+  - components.yaml
+patches:
+  - path: patch.yaml
diff --git a/e2e/manifests/metrics-server/patch.yaml b/e2e/manifests/metrics-server/patch.yaml
@@ -0,0 +1,20 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  labels:
+    k8s-app: metrics-server
+  name: metrics-server
+  namespace: kube-system
+spec:
+  template:
+    spec:
+      containers:
+      - args:
+        - --cert-dir=/tmp
+        - --secure-port=10250
+        - --kubelet-preferred-address-types=InternalIP,ExternalIP,Hostname
+        - --kubelet-use-node-status-port
+        - --kubelet-insecure-tls
+        - --metric-resolution=15s
+        name: metrics-server
+
diff --git a/e2e/manifests/podmetrics.yaml b/e2e/manifests/podmetrics.yaml
@@ -0,0 +1 @@
+
diff --git a/e2e/zombie-detector_test.go b/e2e/zombie-detector_test.go
@@ -2,6 +2,7 @@ package e2e
 
 import (
 	"fmt"
+	"strings"
 	"time"
 
 	. "github.com/onsi/ginkgo/v2"
@@ -26,6 +27,9 @@ var _ = Describe("zombie-detector e2e test", func() {
 		By("waiting for job to be completed")
 		Eventually(func() error {
 			res, err := kubectl(nil, "get", "job", "zombie-detector-immediate-job", "-n", "zombie-detector", "-o", "json")
+			if err != nil {
+				return err
+			}
 			job := batchv1.Job{}
 			err = json.Unmarshal(res, &job)
 			if err != nil {
@@ -42,6 +46,26 @@ var _ = Describe("zombie-detector e2e test", func() {
 		Expect(res.Data).To(BeEmpty())
 	})
 
+	It("should skipping ignored resources", func() {
+		By("creating job from cronjob")
+		_, err := kubectl(nil, "create", "job", "zombie-detector-immediate-job", "-n", "zombie-detector", "--from=cronjob/zombie-detector-cronjob")
+		Expect(err).NotTo(HaveOccurred())
+		By("waiting for job to be completed")
+		Eventually(func() error {
+			res, err := kubectl(nil, "logs", "job/zombie-detector-immediate-job", "-n", "zombie-detector")
+			if err != nil {
+				return err
+			}
+			if !strings.Contains(string(res), "ignoring metrics.k8s.io v1beta1 pods") {
+				return fmt.Errorf("metrics.k8s.io v1beta1 pods is not ignored")
+			}
+			if !strings.Contains(string(res), "ignoring metrics.k8s.io v1beta1 nodes") {
+				return fmt.Errorf("metrics.k8s.io v1beta1 nodes is not ignored")
+			}
+			return nil
+		}).Should(Succeed())
+	})
+
 	It("should detect zombie resources", func() {
 		By("adding deletionTimestamp to resources")
 		_, err := kubectl(nil, "delete", "deployment", "test-deployment", "-n", "default", "--wait=false")
@@ -108,7 +132,7 @@ var _ = Describe("zombie-detector e2e test", func() {
 		Expect(err).NotTo(HaveOccurred())
 		index, err := returnZombieDetectorMetricsIndex(*res)
 		Expect(err).NotTo(HaveOccurred())
-		Expect(len(res.Data[index].ZombieDurationHours.Metrics)).To(Equal(3))
+		Expect(len(res.Data[index].ZombieDurationSeconds.Metrics)).To(Equal(3))
 	})
 
 	It("should not detect anything again", func() {
@@ -196,7 +220,7 @@ type Response struct {
 		Labels struct {
 			Job string `json:"job"`
 		} `json:"labels"`
-		ZombieDurationHours struct {
+		ZombieDurationSeconds struct {
 			Metrics []struct {
 				Labels struct {
 					APIVersion string    `json:"apiVersion"`
@@ -209,6 +233,6 @@ type Response struct {
 				} `json:"labels"`
 				Value string `json:"value"`
 			} `json:"metrics"`
-		} `json:"zombie_duration_hours,omitempty"`
+		} `json:"zombie_duration_seconds,omitempty"`
 	} `json:"data"`
 }
diff --git a/go.mod b/go.mod
diff --git a/go.sum b/go.sum

Original file line number	Diff line number	Diff line change
`@@ -0,0 +1,3 @@`
	`1`	`+.PHONY: update-manifests`
	`2`	`+update-manifests:`
	`3`	`+ https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml`