send bookmark right now after sending all items in watchCache store

Author: Chaunceyctx

staging/src/k8s.io/apiserver/pkg/storage/cacher/cache_watcher.go

@@ -503,6 +503,10 @@ func (c *cacheWatcher) processInterval(ctx context.Context, cacheInterval *watch
 		klog.V(2).Infof("processing %d initEvents of %s (%s) took %v", initEventCount, c.groupResource, c.identifier, processingTime)
 	}
 
+	// send bookmark after sending all events in cacheInterval for watchlist request
+	if cacheInterval.initialEventsEndBookmark != nil {
+		c.sendWatchCacheEvent(cacheInterval.initialEventsEndBookmark)
+	}
 	c.process(ctx, resourceVersion)
 }
 

staging/src/k8s.io/apiserver/pkg/storage/cacher/cacher.go

@@ -653,6 +653,8 @@ func (c *Cacher) Watch(ctx context.Context, key string, opts storage.ListOptions
 		return newErrWatcher(err), nil
 	}
 
+	c.setInitialEventsEndBookmarkIfRequested(cacheInterval, opts, c.watchCache.resourceVersion)
+
 	addedWatcher := false
 	func() {
 		c.Lock()
@@ -1439,6 +1441,26 @@ func (c *Cacher) Wait(ctx context.Context) error {
 	return c.ready.wait(ctx)
 }
 
+// setInitialEventsEndBookmarkIfRequested sets initialEventsEndBookmark field in watchCacheInterval for watchlist request
+func (c *Cacher) setInitialEventsEndBookmarkIfRequested(cacheInterval *watchCacheInterval, opts storage.ListOptions, currentResourceVersion uint64) {
+	if opts.SendInitialEvents != nil && *opts.SendInitialEvents && opts.Predicate.AllowWatchBookmarks {
+		// We don't need to set the InitialEventsAnnotation for this bookmark event,
+		// because this will be automatically set during event conversion in cacheWatcher.convertToWatchEvent method
+		initialEventsEndBookmark := &watchCacheEvent{
+			Type:            watch.Bookmark,
+			Object:          c.newFunc(),
+			ResourceVersion: currentResourceVersion,
+		}
+
+		if err := c.versioner.UpdateObject(initialEventsEndBookmark.Object, initialEventsEndBookmark.ResourceVersion); err != nil {
+			klog.Errorf("failure to set resourceVersion to %d on initialEventsEndBookmark event %+v for watchlist request and wait for bookmark trigger to send", initialEventsEndBookmark.ResourceVersion, initialEventsEndBookmark.Object)
+			initialEventsEndBookmark = nil
+		}
+
+		cacheInterval.initialEventsEndBookmark = initialEventsEndBookmark
+	}
+}
+
 // errWatcher implements watch.Interface to return a single error
 type errWatcher struct {
 	result chan watch.Event

staging/src/k8s.io/apiserver/pkg/storage/cacher/cacher_whitebox_test.go

@@ -49,6 +49,7 @@ import (
 	"k8s.io/apiserver/pkg/storage/cacher/metrics"
 	etcd3testing "k8s.io/apiserver/pkg/storage/etcd3/testing"
 	etcdfeature "k8s.io/apiserver/pkg/storage/feature"
+	storagetesting "k8s.io/apiserver/pkg/storage/testing"
 	utilfeature "k8s.io/apiserver/pkg/util/feature"
 	featuregatetesting "k8s.io/component-base/featuregate/testing"
 	k8smetrics "k8s.io/component-base/metrics"
@@ -1171,6 +1172,106 @@ func TestCacherSendBookmarkEvents(t *testing.T) {
 	}
 }
 
+func TestInitialEventsEndBookmark(t *testing.T) {
+	featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.WatchList, true)
+	forceRequestWatchProgressSupport(t)
+
+	backingStorage := &dummyStorage{}
+	cacher, _, err := newTestCacher(backingStorage)
+	if err != nil {
+		t.Fatalf("Couldn't create cacher: %v", err)
+	}
+	defer cacher.Stop()
+
+	if !utilfeature.DefaultFeatureGate.Enabled(features.ResilientWatchCacheInitialization) {
+		if err := cacher.ready.wait(context.Background()); err != nil {
+			t.Fatalf("unexpected error waiting for the cache to be ready")
+		}
+	}
+
+	makePod := func(index uint64) *example.Pod {
+		return &example.Pod{
+			ObjectMeta: metav1.ObjectMeta{
+				Name:            fmt.Sprintf("pod-%d", index),
+				Namespace:       "ns",
+				ResourceVersion: fmt.Sprintf("%v", 100+index),
+			},
+		}
+	}
+
+	numberOfPods := 3
+	var expectedPodEvents []watch.Event
+	for i := 1; i <= numberOfPods; i++ {
+		pod := makePod(uint64(i))
+		if err := cacher.watchCache.Add(pod); err != nil {
+			t.Fatalf("failed to add a pod: %v", err)
+		}
+		expectedPodEvents = append(expectedPodEvents, watch.Event{Type: watch.Added, Object: pod})
+	}
+	var currentResourceVersion uint64 = 100 + 3
+
+	trueVal, falseVal := true, false
+
+	scenarios := []struct {
+		name                string
+		allowWatchBookmarks bool
+		sendInitialEvents   *bool
+	}{
+		{
+			name:                "allowWatchBookmarks=false, sendInitialEvents=false",
+			allowWatchBookmarks: false,
+			sendInitialEvents:   &falseVal,
+		},
+		{
+			name:                "allowWatchBookmarks=false, sendInitialEvents=true",
+			allowWatchBookmarks: false,
+			sendInitialEvents:   &trueVal,
+		},
+		{
+			name:                "allowWatchBookmarks=true, sendInitialEvents=true",
+			allowWatchBookmarks: true,
+			sendInitialEvents:   &trueVal,
+		},
+		{
+			name:                "allowWatchBookmarks=true, sendInitialEvents=false",
+			allowWatchBookmarks: true,
+			sendInitialEvents:   &falseVal,
+		},
+		{
+			name:                "allowWatchBookmarks=false, sendInitialEvents=nil",
+			allowWatchBookmarks: true,
+		},
+	}
+
+	for _, scenario := range scenarios {
+		t.Run(scenario.name, func(t *testing.T) {
+			expectedWatchEvents := expectedPodEvents
+			if scenario.allowWatchBookmarks && scenario.sendInitialEvents != nil && *scenario.sendInitialEvents {
+				expectedWatchEvents = append(expectedWatchEvents, watch.Event{
+					Type: watch.Bookmark,
+					Object: &example.Pod{
+						ObjectMeta: metav1.ObjectMeta{
+							ResourceVersion: strconv.FormatUint(currentResourceVersion, 10),
+							Annotations:     map[string]string{metav1.InitialEventsAnnotationKey: "true"},
+						},
+					},
+				})
+			}
+
+			pred := storage.Everything
+			pred.AllowWatchBookmarks = scenario.allowWatchBookmarks
+			ctx, cancel := context.WithTimeout(context.Background(), 10*time.Minute)
+			defer cancel()
+			w, err := cacher.Watch(ctx, "pods/ns", storage.ListOptions{ResourceVersion: "100", SendInitialEvents: scenario.sendInitialEvents, Predicate: pred})
+			if err != nil {
+				t.Fatalf("Failed to create watch: %v", err)
+			}
+			storagetesting.TestCheckResultsInStrictOrder(t, w, expectedWatchEvents)
+			storagetesting.TestCheckNoMoreResultsWithIgnoreFunc(t, w, nil)
+		})
+	}
+}
+
 func TestCacherSendsMultipleWatchBookmarks(t *testing.T) {
 	backingStorage := &dummyStorage{}
 	cacher, _, err := newTestCacher(backingStorage)

staging/src/k8s.io/apiserver/pkg/storage/cacher/watch_cache_interval.go

@@ -97,6 +97,9 @@ type watchCacheInterval struct {
 	// Given that indexer and indexValidator only read state, if
 	// possible, Locker obtained through RLocker() is provided.
 	lock sync.Locker
+
+	// initialEventsEndBookmark will be sent after sending all events in cacheInterval
+	initialEventsEndBookmark *watchCacheEvent
 }
 
 type attrFunc func(runtime.Object) (labels.Set, fields.Set, error)

staging/src/k8s.io/apiserver/pkg/storage/testing/utils.go

@@ -169,6 +169,36 @@ func testCheckResultFunc(t *testing.T, w watch.Interface, check func(actualEvent
 	}
 }
 
+func testCheckResultWithIgnoreFunc(t *testing.T, w watch.Interface, expectedEvents []watch.Event, ignore func(watch.Event) bool) {
+	checkIndex := 0
+	for {
+		select {
+		case event := <-w.ResultChan():
+			obj := event.Object
+			if co, ok := obj.(runtime.CacheableObject); ok {
+				event.Object = co.GetObject()
+			}
+			if ignore != nil && ignore(event) {
+				continue
+			}
+			if checkIndex < len(expectedEvents) {
+				expectNoDiff(t, "incorrect event", expectedEvents[checkIndex], event)
+				checkIndex++
+			} else {
+				t.Fatalf("cannot receive correct event, expect no event, but get a event: %+v", event)
+			}
+		case <-time.After(100 * time.Millisecond):
+			// wait 100ms forcibly in order to receive watchEvents including bookmark event.
+			// we cannot guarantee that we will receive all bookmark events within 100ms,
+			// but too large timeout value will lead to exceed the timeout of package test.
+			if checkIndex < len(expectedEvents) {
+				t.Fatalf("cannot receive enough events within specific time, rest expected events: %+v", expectedEvents[checkIndex:])
+			}
+			return
+		}
+	}
+}
+
 func testCheckStop(t *testing.T, w watch.Interface) {
 	select {
 	case e, ok := <-w.ResultChan():
@@ -187,16 +217,18 @@ func testCheckStop(t *testing.T, w watch.Interface) {
 	}
 }
 
-func testCheckResultsInStrictOrder(t *testing.T, w watch.Interface, expectedEvents []watch.Event) {
+func TestCheckResultsInStrictOrder(t *testing.T, w watch.Interface, expectedEvents []watch.Event) {
 	for _, expectedEvent := range expectedEvents {
 		testCheckResult(t, w, expectedEvent)
 	}
 }
 
-func testCheckNoMoreResults(t *testing.T, w watch.Interface) {
+func TestCheckNoMoreResultsWithIgnoreFunc(t *testing.T, w watch.Interface, ignore func(watch.Event) bool) {
 	select {
 	case e := <-w.ResultChan():
-		t.Errorf("Unexpected: %#v event received, expected no events", e)
+		if ignore == nil || !ignore(e) {
+			t.Errorf("Unexpected: %#v event received, expected no events", e)
+		}
 	// We consciously make the timeout short here to speed up tests.
 	case <-time.After(100 * time.Millisecond):
 		return

staging/src/k8s.io/apiserver/pkg/storage/testing/watcher_tests.go

@@ -1478,7 +1478,7 @@ func RunWatchSemantics(ctx context.Context, t *testing.T, store storage.Interfac
 			defer w.Stop()
 
 			// make sure we only get initial events
-			testCheckResultsInStrictOrder(t, w, scenario.expectedInitialEvents(createdPods))
+			TestCheckResultsInStrictOrder(t, w, scenario.expectedInitialEvents(createdPods))
 
 			// make sure that the actual bookmark has at least RV >= to the expected one
 			if scenario.expectedInitialEventsBookmarkWithMinimalRV != nil {
@@ -1512,8 +1512,9 @@ func RunWatchSemantics(ctx context.Context, t *testing.T, store storage.Interfac
 				require.NoError(t, err, "failed to add a pod: %v")
 				createdPods = append(createdPods, out)
 			}
-			testCheckResultsInStrictOrder(t, w, scenario.expectedEventsAfterEstablishingWatch(createdPods))
-			testCheckNoMoreResults(t, w)
+			ignoreEventsFn := func(event watch.Event) bool { return event.Type == watch.Bookmark }
+			testCheckResultWithIgnoreFunc(t, w, scenario.expectedEventsAfterEstablishingWatch(createdPods), ignoreEventsFn)
+			TestCheckNoMoreResultsWithIgnoreFunc(t, w, ignoreEventsFn)
 		})
 	}
 }
@@ -1567,8 +1568,8 @@ func RunWatchSemanticInitialEventsExtended(ctx context.Context, t *testing.T, st
 
 	// make sure we only get initial events from the first ns
 	// followed by the bookmark with the global RV
-	testCheckResultsInStrictOrder(t, w, expectedInitialEventsInStrictOrder(initialPods, otherNsPod.ResourceVersion))
-	testCheckNoMoreResults(t, w)
+	TestCheckResultsInStrictOrder(t, w, expectedInitialEventsInStrictOrder(initialPods, otherNsPod.ResourceVersion))
+	TestCheckNoMoreResultsWithIgnoreFunc(t, w, nil)
 }
 
 func RunWatchListMatchSingle(ctx context.Context, t *testing.T, store storage.Interface) {
@@ -1622,8 +1623,8 @@ func RunWatchListMatchSingle(ctx context.Context, t *testing.T, store storage.In
 
 	// make sure we only get a single pod matching the field selector
 	// followed by the bookmark with the global RV
-	testCheckResultsInStrictOrder(t, w, expectedInitialEventsInStrictOrder(expectedPod, lastAddedPod.ResourceVersion))
-	testCheckNoMoreResults(t, w)
+	TestCheckResultsInStrictOrder(t, w, expectedInitialEventsInStrictOrder(expectedPod, lastAddedPod.ResourceVersion))
+	TestCheckNoMoreResultsWithIgnoreFunc(t, w, nil)
 }
 
 func makePod(namePrefix string) *example.Pod {