Merge pull request #17133 from derekwaynecarr/quota_controller_improvements

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

@@ -36,6 +36,7 @@ import (
 	client "k8s.io/kubernetes/pkg/client/unversioned"
 	"k8s.io/kubernetes/pkg/client/unversioned/clientcmd"
 	"k8s.io/kubernetes/pkg/cloudprovider"
+	"k8s.io/kubernetes/pkg/controller"
 	"k8s.io/kubernetes/pkg/controller/daemon"
 	"k8s.io/kubernetes/pkg/controller/deployment"
 	endpointcontroller "k8s.io/kubernetes/pkg/controller/endpoint"
@@ -72,6 +73,7 @@ type CMServer struct {
 	ConcurrentRCSyncs                 int
 	ConcurrentDSCSyncs                int
 	ConcurrentJobSyncs                int
+	ConcurrentResourceQuotaSyncs      int
 	ServiceSyncPeriod                 time.Duration
 	NodeSyncPeriod                    time.Duration
 	ResourceQuotaSyncPeriod           time.Duration
@@ -113,9 +115,10 @@ func NewCMServer() *CMServer {
 		ConcurrentRCSyncs:                 5,
 		ConcurrentDSCSyncs:                2,
 		ConcurrentJobSyncs:                5,
+		ConcurrentResourceQuotaSyncs:      5,
 		ServiceSyncPeriod:                 5 * time.Minute,
 		NodeSyncPeriod:                    10 * time.Second,
-		ResourceQuotaSyncPeriod:           10 * time.Second,
+		ResourceQuotaSyncPeriod:           5 * time.Minute,
 		NamespaceSyncPeriod:               5 * time.Minute,
 		PVClaimBinderSyncPeriod:           10 * time.Minute,
 		HorizontalPodAutoscalerSyncPeriod: 30 * time.Second,
@@ -183,6 +186,7 @@ func (s *CMServer) AddFlags(fs *pflag.FlagSet) {
 	fs.StringVar(&s.CloudConfigFile, "cloud-config", s.CloudConfigFile, "The path to the cloud provider configuration file.  Empty string for no configuration file.")
 	fs.IntVar(&s.ConcurrentEndpointSyncs, "concurrent-endpoint-syncs", s.ConcurrentEndpointSyncs, "The number of endpoint syncing operations that will be done concurrently. Larger number = faster endpoint updating, but more CPU (and network) load")
 	fs.IntVar(&s.ConcurrentRCSyncs, "concurrent_rc_syncs", s.ConcurrentRCSyncs, "The number of replication controllers that are allowed to sync concurrently. Larger number = more reponsive replica management, but more CPU (and network) load")
+	fs.IntVar(&s.ConcurrentResourceQuotaSyncs, "concurrent-resource-quota-syncs", s.ConcurrentResourceQuotaSyncs, "The number of resource quotas that are allowed to sync concurrently. Larger number = more responsive quota management, but more CPU (and network) load")
 	fs.DurationVar(&s.ServiceSyncPeriod, "service-sync-period", s.ServiceSyncPeriod, "The period for syncing services with their external load balancers")
 	fs.DurationVar(&s.NodeSyncPeriod, "node-sync-period", s.NodeSyncPeriod, ""+
 		"The period for syncing nodes from cloudprovider. Longer periods will result in "+
@@ -304,7 +308,8 @@ func (s *CMServer) Run(_ []string) error {
 		glog.Infof("allocate-node-cidrs set to %v, node controller not creating routes", s.AllocateNodeCIDRs)
 	}
 
-	resourcequotacontroller.NewResourceQuotaController(kubeClient).Run(s.ResourceQuotaSyncPeriod)
+	go resourcequotacontroller.NewResourceQuotaController(
+		kubeClient, controller.StaticResyncPeriodFunc(s.ResourceQuotaSyncPeriod)).Run(s.ConcurrentResourceQuotaSyncs, util.NeverStop)
 
 	// If apiserver is not running we should wait for some time and fail only then. This is particularly
 	// important when we start apiserver and controller manager at the same time.

contrib/mesos/pkg/controllermanager/controllermanager.go

@@ -33,6 +33,7 @@ import (
 	clientcmdapi "k8s.io/kubernetes/pkg/client/unversioned/clientcmd/api"
 	"k8s.io/kubernetes/pkg/cloudprovider"
 	"k8s.io/kubernetes/pkg/cloudprovider/providers/mesos"
+	"k8s.io/kubernetes/pkg/controller"
 	"k8s.io/kubernetes/pkg/controller/daemon"
 	kendpoint "k8s.io/kubernetes/pkg/controller/endpoint"
 	namespacecontroller "k8s.io/kubernetes/pkg/controller/namespace"
@@ -160,8 +161,8 @@ func (s *CMServer) Run(_ []string) error {
 		routeController.Run(s.NodeSyncPeriod)
 	}
 
-	resourceQuotaController := resourcequotacontroller.NewResourceQuotaController(kubeClient)
-	resourceQuotaController.Run(s.ResourceQuotaSyncPeriod)
+	go resourcequotacontroller.NewResourceQuotaController(
+		kubeClient, controller.StaticResyncPeriodFunc(s.ResourceQuotaSyncPeriod)).Run(s.ConcurrentResourceQuotaSyncs, util.NeverStop)
 
 	namespaceController := namespacecontroller.NewNamespaceController(kubeClient, &unversioned.APIVersions{}, s.NamespaceSyncPeriod)
 	namespaceController.Run()

docs/admin/kube-controller-manager.md

@@ -61,6 +61,7 @@ kube-controller-manager
       --cluster-cidr=<nil>: CIDR Range for Pods in cluster.
       --cluster-name="kubernetes": The instance prefix for the cluster
       --concurrent-endpoint-syncs=5: The number of endpoint syncing operations that will be done concurrently. Larger number = faster endpoint updating, but more CPU (and network) load
+      --concurrent-resource-quota-syncs=5: The number of resource quotas that are allowed to sync concurrently. Larger number = more responsive quota management, but more CPU (and network) load
       --concurrent_rc_syncs=5: The number of replication controllers that are allowed to sync concurrently. Larger number = more reponsive replica management, but more CPU (and network) load
       --deleting-pods-burst=10: Number of nodes on which pods are bursty deleted in case of node failure. For more details look into RateLimiter.
       --deleting-pods-qps=0.1: Number of nodes per second on which pods are deleted in case of node failure.
@@ -88,14 +89,14 @@ kube-controller-manager
       --pv-recycler-pod-template-filepath-nfs="": The file path to a pod definition used as a template for NFS persistent volume recycling
       --pv-recycler-timeout-increment-hostpath=30: the increment of time added per Gi to ActiveDeadlineSeconds for a HostPath scrubber pod.  This is for development and testing only and will not work in a multi-node cluster.
       --pvclaimbinder-sync-period=10m0s: The period for syncing persistent volumes and persistent volume claims
-      --resource-quota-sync-period=10s: The period for syncing quota usage status in the system
+      --resource-quota-sync-period=5m0s: The period for syncing quota usage status in the system
       --root-ca-file="": If set, this root certificate authority will be included in service account's token secret. This must be a valid PEM-encoded CA bundle.
       --service-account-private-key-file="": Filename containing a PEM-encoded private RSA key used to sign service account tokens.
       --service-sync-period=5m0s: The period for syncing services with their external load balancers
       --terminated-pod-gc-threshold=12500: Number of terminated pods that can exist before the terminated pod garbage collector starts deleting terminated pods. If <= 0, the terminated pod garbage collector is disabled.
 ```
 
-###### Auto generated by spf13/cobra on 4-Nov-2015
+###### Auto generated by spf13/cobra on 30-Nov-2015
 
 
 <!-- BEGIN MUNGE: GENERATED_ANALYTICS -->

hack/verify-flags/known-flags.txt

@@ -46,6 +46,7 @@ cluster-domain
 cluster-name
 cluster-tag
 concurrent-endpoint-syncs
+concurrent-resource-quota-syncs
 config-sync-period
 configure-cbr0
 container-port

pkg/client/cache/listers.go

@@ -336,7 +336,7 @@ func (s *StoreToJobLister) List() (jobs extensions.JobList, err error) {
 	return jobs, nil
 }
 
-// GetPodControllers returns a list of jobs managing a pod. Returns an error only if no matching jobs are found.
+// GetPodJobs returns a list of jobs managing a pod. Returns an error only if no matching jobs are found.
 func (s *StoreToJobLister) GetPodJobs(pod *api.Pod) (jobs []extensions.Job, err error) {
 	var selector labels.Selector
 	var job extensions.Job

pkg/controller/controller_utils.go

@@ -61,6 +61,13 @@ func NoResyncPeriodFunc() time.Duration {
 	return 0
 }
 
+// StaticResyncPeriodFunc returns the resync period specified
+func StaticResyncPeriodFunc(resyncPeriod time.Duration) ResyncPeriodFunc {
+	return func() time.Duration {
+		return resyncPeriod
+	}
+}
+
 // Expectations are a way for controllers to tell the controller manager what they expect. eg:
 //	ControllerExpectations: {
 //		controller1: expects  2 adds in 2 minutes

pkg/controller/resourcequota/resource_quota_controller_test.go

@@ -19,10 +19,12 @@ package resourcequota
 import (
 	"strconv"
 	"testing"
+	"time"
 
 	"k8s.io/kubernetes/pkg/api"
 	"k8s.io/kubernetes/pkg/api/resource"
 	"k8s.io/kubernetes/pkg/client/unversioned/testclient"
+	"k8s.io/kubernetes/pkg/controller"
 	"k8s.io/kubernetes/pkg/util/sets"
 )
 
@@ -173,7 +175,7 @@ func TestSyncResourceQuota(t *testing.T) {
 
 	kubeClient := testclient.NewSimpleFake(&podList, &quota)
 
-	ResourceQuotaController := NewResourceQuotaController(kubeClient)
+	ResourceQuotaController := NewResourceQuotaController(kubeClient, controller.StaticResyncPeriodFunc(time.Second))
 	err := ResourceQuotaController.syncResourceQuota(quota)
 	if err != nil {
 		t.Fatalf("Unexpected error %v", err)
@@ -230,7 +232,7 @@ func TestSyncResourceQuotaSpecChange(t *testing.T) {
 
 	kubeClient := testclient.NewSimpleFake(&quota)
 
-	ResourceQuotaController := NewResourceQuotaController(kubeClient)
+	ResourceQuotaController := NewResourceQuotaController(kubeClient, controller.StaticResyncPeriodFunc(time.Second))
 	err := ResourceQuotaController.syncResourceQuota(quota)
 	if err != nil {
 		t.Fatalf("Unexpected error %v", err)
@@ -277,7 +279,7 @@ func TestSyncResourceQuotaNoChange(t *testing.T) {
 
 	kubeClient := testclient.NewSimpleFake(&api.PodList{}, &quota)
 
-	ResourceQuotaController := NewResourceQuotaController(kubeClient)
+	ResourceQuotaController := NewResourceQuotaController(kubeClient, controller.StaticResyncPeriodFunc(time.Second))
 	err := ResourceQuotaController.syncResourceQuota(quota)
 	if err != nil {
 		t.Fatalf("Unexpected error %v", err)