Merge pull request kubernetes#119185 from xing-yang/metrics_attach

Add reason to force detach metric

Author: k8s-ci-robot

pkg/controller/volume/attachdetach/metrics/metrics.go

@@ -35,6 +35,14 @@ import (
 
 const pluginNameNotAvailable = "N/A"
 
+const (
+	// Force detach reason is timeout
+	ForceDetachReasonTimeout = "timeout"
+	// Force detach reason is the node has an out-of-service taint
+	ForceDetachReasonOutOfService = "out-of-service"
+	attachDetachController        = "attach_detach_controller"
+)
+
 var (
 	inUseVolumeMetricDesc = metrics.NewDesc(
 		metrics.BuildFQName("", "storage_count", "attachable_volumes_in_use"),
@@ -48,12 +56,15 @@ var (
 		[]string{"plugin_name", "state"}, nil,
 		metrics.ALPHA, "")
 
-	forcedDetachMetricCounter = metrics.NewCounter(
+	ForceDetachMetricCounter = metrics.NewCounterVec(
 		&metrics.CounterOpts{
+			Subsystem:      attachDetachController,
 			Name:           "attachdetach_controller_forced_detaches",
 			Help:           "Number of times the A/D Controller performed a forced detach",
 			StabilityLevel: metrics.ALPHA,
-		})
+		},
+		[]string{"reason"},
+	)
 )
 var registerMetrics sync.Once
 
@@ -75,7 +86,7 @@ func Register(pvcLister corelisters.PersistentVolumeClaimLister,
 			pluginMgr,
 			csiMigratedPluginManager,
 			intreeToCSITranslator))
-		legacyregistry.MustRegister(forcedDetachMetricCounter)
+		legacyregistry.MustRegister(ForceDetachMetricCounter)
 	})
 }
 
@@ -209,6 +220,6 @@ func (collector *attachDetachStateCollector) getTotalVolumesCount() volumeCount
 }
 
 // RecordForcedDetachMetric register a forced detach metric.
-func RecordForcedDetachMetric() {
-	forcedDetachMetricCounter.Inc()
+func RecordForcedDetachMetric(forceDetachReason string) {
+	ForceDetachMetricCounter.WithLabelValues(forceDetachReason).Inc()
 }

pkg/controller/volume/attachdetach/reconciler/reconciler.go

@@ -269,14 +269,21 @@ func (rc *reconciler) reconcile(ctx context.Context) {
 			verifySafeToDetach := !(timeout || hasOutOfServiceTaint)
 			err = rc.attacherDetacher.DetachVolume(logger, attachedVolume.AttachedVolume, verifySafeToDetach, rc.actualStateOfWorld)
 			if err == nil {
-				if !timeout {
+				if verifySafeToDetach { // normal detach
 					logger.Info("attacherDetacher.DetachVolume started", "node", klog.KRef("", string(attachedVolume.NodeName)), "volumeName", attachedVolume.VolumeName)
-				} else {
-					metrics.RecordForcedDetachMetric()
-					logger.Info("attacherDetacher.DetachVolume started: this volume is not safe to detach, but maxWaitForUnmountDuration expired, force detaching",
-						"duration", rc.maxWaitForUnmountDuration,
-						"node", klog.KRef("", string(attachedVolume.NodeName)),
-						"volumeName", attachedVolume.VolumeName)
+				} else { // force detach
+					if timeout {
+						metrics.RecordForcedDetachMetric(metrics.ForceDetachReasonTimeout)
+						logger.Info("attacherDetacher.DetachVolume started: this volume is not safe to detach, but maxWaitForUnmountDuration expired, force detaching",
+							"duration", rc.maxWaitForUnmountDuration,
+							"node", klog.KRef("", string(attachedVolume.NodeName)),
+							"volumeName", attachedVolume.VolumeName)
+					} else {
+						metrics.RecordForcedDetachMetric(metrics.ForceDetachReasonOutOfService)
+						logger.Info("attacherDetacher.DetachVolume started: node has out-of-service taint, force detaching",
+							"node", klog.KRef("", string(attachedVolume.NodeName)),
+							"volumeName", attachedVolume.VolumeName)
+					}
 				}
 			}
 			if err != nil {

pkg/controller/volume/attachdetach/reconciler/reconciler_test.go

@@ -19,6 +19,7 @@ package reconciler
 import (
 	"context"
 	"fmt"
+	"sync"
 	"testing"
 	"time"
 
@@ -30,10 +31,13 @@ import (
 	"k8s.io/client-go/informers"
 	"k8s.io/client-go/tools/record"
 	featuregatetesting "k8s.io/component-base/featuregate/testing"
+	"k8s.io/component-base/metrics/legacyregistry"
+	metricstestutil "k8s.io/component-base/metrics/testutil"
 	"k8s.io/klog/v2"
 	"k8s.io/klog/v2/ktesting"
 	"k8s.io/kubernetes/pkg/controller"
 	"k8s.io/kubernetes/pkg/controller/volume/attachdetach/cache"
+	"k8s.io/kubernetes/pkg/controller/volume/attachdetach/metrics"
 	"k8s.io/kubernetes/pkg/controller/volume/attachdetach/statusupdater"
 	controllervolumetesting "k8s.io/kubernetes/pkg/controller/volume/attachdetach/testing"
 	"k8s.io/kubernetes/pkg/features"
@@ -51,6 +55,8 @@ const (
 	volumeAttachedCheckTimeout    = 5 * time.Second
 )
 
+var registerMetrics sync.Once
+
 // Calls Run()
 // Verifies there are no calls to attach or detach.
 func Test_Run_Positive_DoNothing(t *testing.T) {
@@ -221,6 +227,9 @@ func Test_Run_Positive_OneDesiredVolumeAttachThenDetachWithUnmountedVolume(t *te
 // Deletes the node/volume/pod tuple from desiredStateOfWorld cache without first marking the node/volume as unmounted.
 // Verifies there is one detach call and no (new) attach calls.
 func Test_Run_Positive_OneDesiredVolumeAttachThenDetachWithMountedVolume(t *testing.T) {
+	registerMetrics.Do(func() {
+		legacyregistry.MustRegister(metrics.ForceDetachMetricCounter)
+	})
 	// Arrange
 	volumePluginMgr, fakePlugin := volumetesting.GetTestVolumePluginMgr(t)
 	dsw := cache.NewDesiredStateOfWorld(volumePluginMgr)
@@ -287,6 +296,9 @@ func Test_Run_Positive_OneDesiredVolumeAttachThenDetachWithMountedVolume(t *test
 	waitForAttachCallCount(t, 1 /* expectedAttachCallCount */, fakePlugin)
 	verifyNewDetacherCallCount(t, false /* expectZeroNewDetacherCallCount */, fakePlugin)
 	waitForDetachCallCount(t, 1 /* expectedDetachCallCount */, fakePlugin)
+
+	// Force detach metric due to timeout
+	testForceDetachMetric(t, 1, metrics.ForceDetachReasonTimeout)
 }
 
 // Populates desiredStateOfWorld cache with one node/volume/pod tuple.
@@ -852,6 +864,9 @@ func Test_Run_OneVolumeAttachAndDetachTimeoutNodesWithReadWriteOnce(t *testing.T
 // Verifies there is one detach call and no (new) attach calls.
 func Test_Run_OneVolumeDetachOnOutOfServiceTaintedNode(t *testing.T) {
 	defer featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.NodeOutOfServiceVolumeDetach, true)()
+	registerMetrics.Do(func() {
+		legacyregistry.MustRegister(metrics.ForceDetachMetricCounter)
+	})
 	// Arrange
 	volumePluginMgr, fakePlugin := volumetesting.GetTestVolumePluginMgr(t)
 	dsw := cache.NewDesiredStateOfWorld(volumePluginMgr)
@@ -920,6 +935,9 @@ func Test_Run_OneVolumeDetachOnOutOfServiceTaintedNode(t *testing.T) {
 	waitForAttachCallCount(t, 1 /* expectedAttachCallCount */, fakePlugin)
 	verifyNewDetacherCallCount(t, false /* expectZeroNewDetacherCallCount */, fakePlugin)
 	waitForDetachCallCount(t, 1 /* expectedDetachCallCount */, fakePlugin)
+
+	// Force detach metric due to out-of-service taint
+	testForceDetachMetric(t, 1, metrics.ForceDetachReasonOutOfService)
 }
 
 // Populates desiredStateOfWorld cache with one node/volume/pod tuple.
@@ -1666,3 +1684,16 @@ func retryWithExponentialBackOff(initialDuration time.Duration, fn wait.Conditio
 	}
 	return wait.ExponentialBackoff(backoff, fn)
 }
+
+// verifies the force detach metric with reason
+func testForceDetachMetric(t *testing.T, inputForceDetachMetricCounter int, reason string) {
+	t.Helper()
+
+	actualForceDetachMericCounter, err := metricstestutil.GetCounterMetricValue(metrics.ForceDetachMetricCounter.WithLabelValues(reason))
+	if err != nil {
+		t.Errorf("Error getting actualForceDetachMericCounter")
+	}
+	if actualForceDetachMericCounter != float64(inputForceDetachMetricCounter) {
+		t.Errorf("Expected desiredForceDetachMericCounter to be %d, got %v", inputForceDetachMetricCounter, actualForceDetachMericCounter)
+	}
+}