Improve GC discovery sync performance

Improve GC discovery sync performance by only syncing when discovered
resource diffs are detected. Before, the GC worker pool was shut down
and monitors resynced unconditionally every sync period, leading to
significant processing delays causing test flakes where otherwise
reasonable GC timeouts were being exceeded.

Related to https://github.com/kubernetes/kubernetes/issues/49966.

cmd/kube-controller-manager/app/core.go

@@ -339,7 +339,7 @@ func startGarbageCollectorController(ctx ControllerContext) (bool, error) {
 
 	// Periodically refresh the RESTMapper with new discovery information and sync
 	// the garbage collector.
-	go garbageCollector.Sync(restMapper, discoveryClient, 30*time.Second, ctx.Stop)
+	go garbageCollector.Sync(gcClientset.Discovery(), 30*time.Second, ctx.Stop)
 
 	return true, nil
 }

pkg/controller/garbagecollector/garbagecollector.go

@@ -18,6 +18,7 @@ package garbagecollector
 
 import (
 	"fmt"
+	"reflect"
 	"sync"
 	"time"
 
@@ -59,7 +60,7 @@ const ResourceResyncTime time.Duration = 0
 // ensures that the garbage collector operates with a graph that is at least as
 // up to date as the notification is sent.
 type GarbageCollector struct {
-	restMapper meta.RESTMapper
+	restMapper resettableRESTMapper
 	// clientPool uses the regular dynamicCodec. We need it to update
 	// finalizers. It can be removed if we support patching finalizers.
 	clientPool dynamic.ClientPool
@@ -81,7 +82,7 @@ type GarbageCollector struct {
 func NewGarbageCollector(
 	metaOnlyClientPool dynamic.ClientPool,
 	clientPool dynamic.ClientPool,
-	mapper meta.RESTMapper,
+	mapper resettableRESTMapper,
 	deletableResources map[schema.GroupVersionResource]struct{},
 	ignoredResources map[schema.GroupResource]struct{},
 	sharedInformers informers.SharedInformerFactory,
@@ -162,25 +163,59 @@ type resettableRESTMapper interface {
 	Reset()
 }
 
-// Sync periodically resyncs the garbage collector monitors with resources
+// Sync periodically resyncs the garbage collector when new resources are
-// returned found via the discoveryClient. Sync blocks, continuing to sync until
+// observed from discovery. When new resources are detected, Sync will stop all
-// a message is received on stopCh.
+// GC workers, reset gc.restMapper, and resync the monitors.
 //
-// The discoveryClient should be the same client which underlies restMapper.
+// Note that discoveryClient should NOT be shared with gc.restMapper, otherwise
-func (gc *GarbageCollector) Sync(restMapper resettableRESTMapper, discoveryClient discovery.DiscoveryInterface, period time.Duration, stopCh <-chan struct{}) {
+// the mapper's underlying discovery client will be unnecessarily reset during
+// the course of detecting new resources.
+func (gc *GarbageCollector) Sync(discoveryClient discovery.DiscoveryInterface, period time.Duration, stopCh <-chan struct{}) {
+	oldResources := make(map[schema.GroupVersionResource]struct{})
 	wait.Until(func() {
+		// Get the current resource list from discovery.
+		newResources, err := GetDeletableResources(discoveryClient)
+		if err != nil {
+			utilruntime.HandleError(err)
+			return
+		}
+
+		// Detect first or abnormal sync and try again later.
+		if oldResources == nil || len(oldResources) == 0 {
+			oldResources = newResources
+			return
+		}
+
+		// Decide whether discovery has reported a change.
+		if reflect.DeepEqual(oldResources, newResources) {
+			glog.V(5).Infof("no resource updates from discovery, skipping garbage collector sync")
+			return
+		}
+
+		// Something has changed, so track the new state and perform a sync.
+		glog.V(2).Infof("syncing garbage collector with updated resources from discovery: %v", newResources)
+		oldResources = newResources
+
 		// Ensure workers are paused to avoid processing events before informers
 		// have resynced.
 		gc.workerLock.Lock()
 		defer gc.workerLock.Unlock()
 
-		restMapper.Reset()
+		// Resetting the REST mapper will also invalidate the underlying discovery
-		deletableResources, err := GetDeletableResources(discoveryClient)
+		// client. This is a leaky abstraction and assumes behavior about the REST
-		if err != nil {
+		// mapper, but we'll deal with it for now.
-			utilruntime.HandleError(err)
+		gc.restMapper.Reset()
-			return
+
-		}
+		// Perform the monitor resync and wait for controllers to report cache sync.
-		if err := gc.resyncMonitors(deletableResources); err != nil {
+		//
+		// NOTE: It's possible that newResources will diverge from the resources
+		// discovered by restMapper during the call to Reset, since they are
+		// distinct discovery clients invalidated at different times. For example,
+		// newResources may contain resources not returned in the restMapper's
+		// discovery call if the resources appeared inbetween the calls. In that
+		// case, the restMapper will fail to map some of newResources until the next
+		// sync period.
+		if err := gc.resyncMonitors(newResources); err != nil {
 			utilruntime.HandleError(fmt.Errorf("failed to sync resource monitors: %v", err))
 			return
 		}

pkg/controller/garbagecollector/garbagecollector_test.go

@@ -46,11 +46,17 @@ import (
 	"k8s.io/kubernetes/pkg/controller/garbagecollector/metaonly"
 )
 
+type testRESTMapper struct {
+	meta.RESTMapper
+}
+
+func (_ *testRESTMapper) Reset() {}
+
 func TestGarbageCollectorConstruction(t *testing.T) {
 	config := &restclient.Config{}
 	config.ContentConfig.NegotiatedSerializer = serializer.DirectCodecFactory{CodecFactory: metaonly.NewMetadataCodecFactory()}
 	tweakableRM := meta.NewDefaultRESTMapper(nil, nil)
-	rm := meta.MultiRESTMapper{tweakableRM, api.Registry.RESTMapper()}
+	rm := &testRESTMapper{meta.MultiRESTMapper{tweakableRM, api.Registry.RESTMapper()}}
 	metaOnlyClientPool := dynamic.NewClientPool(config, rm, dynamic.LegacyAPIPathResolverFunc)
 	config.ContentConfig.NegotiatedSerializer = nil
 	clientPool := dynamic.NewClientPool(config, rm, dynamic.LegacyAPIPathResolverFunc)
@@ -168,7 +174,7 @@ func setupGC(t *testing.T, config *restclient.Config) garbageCollector {
 	podResource := map[schema.GroupVersionResource]struct{}{{Version: "v1", Resource: "pods"}: {}}
 	client := fake.NewSimpleClientset()
 	sharedInformers := informers.NewSharedInformerFactory(client, 0)
-	gc, err := NewGarbageCollector(metaOnlyClientPool, clientPool, api.Registry.RESTMapper(), podResource, ignoredResources, sharedInformers)
+	gc, err := NewGarbageCollector(metaOnlyClientPool, clientPool, &testRESTMapper{api.Registry.RESTMapper()}, podResource, ignoredResources, sharedInformers)
 	if err != nil {
 		t.Fatal(err)
 	}

test/integration/garbagecollector/garbage_collector_test.go

@@ -265,7 +265,7 @@ func setup(t *testing.T, workerCount int) *testContext {
 	syncPeriod := 5 * time.Second
 	startGC := func(workers int) {
 		go gc.Run(workers, stopCh)
-		go gc.Sync(restMapper, discoveryClient, syncPeriod, stopCh)
+		go gc.Sync(clientSet.Discovery(), syncPeriod, stopCh)
 	}
 
 	if workerCount > 0 {