Bootstrap multi-cluster controller

cmd/glbc/main.go

@@ -26,10 +26,12 @@ import (
 
 	crdclient "k8s.io/apiextensions-apiserver/pkg/client/clientset/clientset"
 	"k8s.io/client-go/kubernetes"
+	crclient "k8s.io/cluster-registry/pkg/client/clientset_generated/clientset"
 	"k8s.io/kubernetes/pkg/cloudprovider/providers/gce"
 
 	"k8s.io/ingress-gce/pkg/context"
 	"k8s.io/ingress-gce/pkg/controller"
+	"k8s.io/ingress-gce/pkg/mci"
 	neg "k8s.io/ingress-gce/pkg/neg"
 
 	"k8s.io/ingress-gce/cmd/glbc/app"
@@ -82,6 +84,14 @@ func main() {
 		}
 	}
 
+	var registryClient crclient.Interface
+	if flags.F.MultiCluster {
+		registryClient, err = crclient.NewForConfig(kubeConfig)
+		if err != nil {
+			glog.Fatalf("Failed to create Cluster Registry client: %v", err)
+		}
+	}
+
 	namer, err := app.NewNamer(kubeClient, flags.F.ClusterName, controller.DefaultFirewallName)
 	if err != nil {
 		glog.Fatalf("%v", err)
@@ -96,8 +106,8 @@ func main() {
 
 	enableNEG := cloud.AlphaFeatureGate.Enabled(gce.AlphaFeatureNetworkEndpointGroup)
 	stopCh := make(chan struct{})
-	ctx := context.NewControllerContext(kubeClient, flags.F.WatchNamespace, flags.F.ResyncPeriod, enableNEG)
-	lbc, err := controller.NewLoadBalancerController(kubeClient, stopCh, ctx, clusterManager, enableNEG)
+	ctx := context.NewControllerContext(kubeClient, registryClient, flags.F.WatchNamespace, flags.F.ResyncPeriod, enableNEG)
+	lbc, err := controller.NewLoadBalancerController(ctx, clusterManager, enableNEG, stopCh)
 	if err != nil {
 		glog.Fatalf("Error creating load balancer controller: %v", err)
 	}
@@ -109,11 +119,17 @@ func main() {
 	glog.V(0).Infof("clusterManager initialized")
 
 	if enableNEG {
-		negController, _ := neg.NewController(kubeClient, cloud, ctx, lbc.Translator, namer, flags.F.ResyncPeriod)
+		negController, _ := neg.NewController(cloud, ctx, lbc.Translator, namer, flags.F.ResyncPeriod)
 		go negController.Run(stopCh)
 		glog.V(0).Infof("negController started")
 	}
 
+	if flags.F.MultiCluster {
+		mciController, _ := mci.NewController(ctx, flags.F.ResyncPeriod)
+		go mciController.Run(stopCh)
+		glog.V(0).Infof("Multi-Cluster Ingress Controller started")
+	}
+
 	go app.RunHTTPServer(lbc)
 	go app.RunSIGTERMHandler(lbc, flags.F.DeleteAllOnQuit)
 

pkg/context/context.go

@@ -30,12 +30,13 @@ import (
 	"k8s.io/client-go/tools/cache"
 	"k8s.io/client-go/tools/record"
 	crclient "k8s.io/cluster-registry/pkg/client/clientset_generated/clientset"
+	crinformerv1alpha1 "k8s.io/cluster-registry/pkg/client/informers_generated/externalversions/clusterregistry/v1alpha1"
 )
 
 // ControllerContext holds resources necessary for the general
 // workflow of the controller.
 type ControllerContext struct {
-	kubeClient kubernetes.Interface
+	KubeClient kubernetes.Interface
 
 	IngressInformer  cache.SharedIndexInformer
 	ServiceInformer  cache.SharedIndexInformer
@@ -46,31 +47,36 @@ type ControllerContext struct {
 	// Map of namespace => record.EventRecorder.
 	recorders map[string]record.EventRecorder
 
-	MCContext MultiClusterContext
+	MC MultiClusterContext
 }
 
 // MultiClusterContext holds resource necessary for MCI mode.
 type MultiClusterContext struct {
-	crClient        crclient.Interface
+	RegistryClient  crclient.Interface
 	ClusterInformer cache.SharedIndexInformer
 }
 
 // NewControllerContext returns a new shared set of informers.
-func NewControllerContext(kubeClient kubernetes.Interface, namespace string, resyncPeriod time.Duration, enableEndpointsInformer bool) *ControllerContext {
+func NewControllerContext(kubeClient kubernetes.Interface, registryClient crclient.Interface, namespace string, resyncPeriod time.Duration, enableEndpointsInformer bool) *ControllerContext {
 	newIndexer := func() cache.Indexers {
 		return cache.Indexers{cache.NamespaceIndex: cache.MetaNamespaceIndexFunc}
 	}
+
 	context := &ControllerContext{
-		kubeClient:      kubeClient,
+		KubeClient:      kubeClient,
 		IngressInformer: informerv1beta1.NewIngressInformer(kubeClient, namespace, resyncPeriod, newIndexer()),
 		ServiceInformer: informerv1.NewServiceInformer(kubeClient, namespace, resyncPeriod, newIndexer()),
 		PodInformer:     informerv1.NewPodInformer(kubeClient, namespace, resyncPeriod, newIndexer()),
 		NodeInformer:    informerv1.NewNodeInformer(kubeClient, resyncPeriod, newIndexer()),
+		MC:              MultiClusterContext{RegistryClient: registryClient},
 		recorders:       map[string]record.EventRecorder{},
 	}
 	if enableEndpointsInformer {
 		context.EndpointInformer = informerv1.NewEndpointsInformer(kubeClient, namespace, resyncPeriod, newIndexer())
 	}
+	if context.MC.RegistryClient != nil {
+		context.MC.ClusterInformer = crinformerv1alpha1.NewClusterInformer(registryClient, resyncPeriod, newIndexer())
+	}
 
 	return context
 }
@@ -86,6 +92,9 @@ func (ctx *ControllerContext) HasSynced() bool {
 	if ctx.EndpointInformer != nil {
 		funcs = append(funcs, ctx.EndpointInformer.HasSynced)
 	}
+	if ctx.MC.ClusterInformer != nil {
+		funcs = append(funcs, ctx.MC.ClusterInformer.HasSynced)
+	}
 	for _, f := range funcs {
 		if !f() {
 			return false
@@ -94,6 +103,7 @@ func (ctx *ControllerContext) HasSynced() bool {
 	return true
 }
 
+// Recorder creates an event recorder for the specified namespace.
 func (ctx *ControllerContext) Recorder(ns string) record.EventRecorder {
 	if rec, ok := ctx.recorders[ns]; ok {
 		return rec
@@ -102,7 +112,7 @@ func (ctx *ControllerContext) Recorder(ns string) record.EventRecorder {
 	broadcaster := record.NewBroadcaster()
 	broadcaster.StartLogging(glog.Infof)
 	broadcaster.StartRecordingToSink(&corev1.EventSinkImpl{
-		Interface: ctx.kubeClient.Core().Events(ns),
+		Interface: ctx.KubeClient.Core().Events(ns),
 	})
 	rec := broadcaster.NewRecorder(scheme.Scheme, apiv1.EventSource{Component: "loadbalancer-controller"})
 	ctx.recorders[ns] = rec
@@ -119,4 +129,8 @@ func (ctx *ControllerContext) Start(stopCh chan struct{}) {
 	if ctx.EndpointInformer != nil {
 		go ctx.EndpointInformer.Run(stopCh)
 	}
+
+	if ctx.MC.ClusterInformer != nil {
+		go ctx.MC.ClusterInformer.Run(stopCh)
+	}
 }

pkg/controller/controller.go

@@ -59,8 +59,7 @@ var (
 // LoadBalancerController watches the kubernetes api and adds/removes services
 // from the loadbalancer, via loadBalancerConfig.
 type LoadBalancerController struct {
-	client kubernetes.Interface
-	ctx    *context.ControllerContext
+	ctx *context.ControllerContext
 
 	ingLister  StoreToIngressLister
 	nodeLister cache.Indexer
@@ -88,18 +87,16 @@ type LoadBalancerController struct {
 }
 
 // NewLoadBalancerController creates a controller for gce loadbalancers.
-// - kubeClient: A kubernetes REST client.
 // - clusterManager: A ClusterManager capable of creating all cloud resources
 //	 required for L7 loadbalancing.
 // - resyncPeriod: Watchers relist from the Kubernetes API server this often.
-func NewLoadBalancerController(kubeClient kubernetes.Interface, stopCh chan struct{}, ctx *context.ControllerContext, clusterManager *ClusterManager, negEnabled bool) (*LoadBalancerController, error) {
+func NewLoadBalancerController(ctx *context.ControllerContext, clusterManager *ClusterManager, negEnabled bool, stopCh chan struct{}) (*LoadBalancerController, error) {
 	broadcaster := record.NewBroadcaster()
 	broadcaster.StartLogging(glog.Infof)
 	broadcaster.StartRecordingToSink(&unversionedcore.EventSinkImpl{
-		Interface: kubeClient.Core().Events(""),
+		Interface: ctx.KubeClient.Core().Events(""),
 	})
 	lbc := LoadBalancerController{
-		client:              kubeClient,
 		ctx:                 ctx,
 		ingLister:           StoreToIngressLister{Store: ctx.IngressInformer.GetStore()},
 		nodeLister:          ctx.NodeInformer.GetIndexer(),
@@ -174,7 +171,7 @@ func NewLoadBalancerController(kubeClient kubernetes.Interface, stopCh chan stru
 		ctx.PodInformer.GetIndexer(),
 		endpointIndexer,
 		negEnabled)
-	lbc.tlsLoader = &tls.TLSCertsFromSecretsLoader{Client: lbc.client}
+	lbc.tlsLoader = &tls.TLSCertsFromSecretsLoader{Client: lbc.ctx.KubeClient}
 
 	glog.V(3).Infof("Created new loadbalancer controller")
 
@@ -299,7 +296,7 @@ func (lbc *LoadBalancerController) ensureIngress(key string, ing *extensions.Ing
 		if err = setInstanceGroupsAnnotation(ing.Annotations, igs); err != nil {
 			return err
 		}
-		if err = updateAnnotations(lbc.client, ing.Name, ing.Namespace, ing.Annotations); err != nil {
+		if err = updateAnnotations(lbc.ctx.KubeClient, ing.Name, ing.Namespace, ing.Annotations); err != nil {
 			return err
 		}
 		return nil
@@ -367,7 +364,7 @@ func (lbc *LoadBalancerController) ensureIngress(key string, ing *extensions.Ing
 // updateIngressStatus updates the IP and annotations of a loadbalancer.
 // The annotations are parsed by kubectl describe.
 func (lbc *LoadBalancerController) updateIngressStatus(l7 *loadbalancers.L7, ing *extensions.Ingress) error {
-	ingClient := lbc.client.Extensions().Ingresses(ing.Namespace)
+	ingClient := lbc.ctx.KubeClient.Extensions().Ingresses(ing.Namespace)
 
 	// Update IP through update/status endpoint
 	ip := l7.GetIP()
@@ -395,7 +392,7 @@ func (lbc *LoadBalancerController) updateIngressStatus(l7 *loadbalancers.L7, ing
 		}
 	}
 	annotations := loadbalancers.GetLBAnnotations(l7, currIng.Annotations, lbc.CloudClusterManager.backendPool)
-	if err := updateAnnotations(lbc.client, ing.Name, ing.Namespace, annotations); err != nil {
+	if err := updateAnnotations(lbc.ctx.KubeClient, ing.Name, ing.Namespace, annotations); err != nil {
 		return err
 	}
 	return nil

pkg/controller/controller_test.go

@@ -55,8 +55,8 @@ func defaultBackendName(clusterName string) string {
 func newLoadBalancerController(t *testing.T, cm *fakeClusterManager) *LoadBalancerController {
 	kubeClient := fake.NewSimpleClientset()
 	stopCh := make(chan struct{})
-	ctx := context.NewControllerContext(kubeClient, api_v1.NamespaceAll, 1*time.Second, true)
-	lb, err := NewLoadBalancerController(kubeClient, stopCh, ctx, cm.ClusterManager, true)
+	ctx := context.NewControllerContext(kubeClient, nil, api_v1.NamespaceAll, 1*time.Second, true)
+	lb, err := NewLoadBalancerController(ctx, cm.ClusterManager, true, stopCh)
 	if err != nil {
 		t.Fatalf("%v", err)
 	}
@@ -202,7 +202,7 @@ func addIngress(lbc *LoadBalancerController, ing *extensions.Ingress, pm *nodePo
 			lbc.ctx.ServiceInformer.GetIndexer().Add(svc)
 		}
 	}
-	lbc.client.Extensions().Ingresses(ing.Namespace).Create(ing)
+	lbc.ctx.KubeClient.Extensions().Ingresses(ing.Namespace).Create(ing)
 	lbc.ctx.IngressInformer.GetIndexer().Add(ing)
 }
 

pkg/controller/translator/translator_test.go

@@ -52,7 +52,7 @@ func gceForTest(negEnabled bool) *GCE {
 
 	namer := utils.NewNamer("uid1", "fw1")
 
-	ctx := context.NewControllerContext(client, apiv1.NamespaceAll, 1*time.Second, negEnabled)
+	ctx := context.NewControllerContext(client, nil, apiv1.NamespaceAll, 1*time.Second, negEnabled)
 	gce := &GCE{
 		recorders:  ctx,
 		namer:      namer,

pkg/mci/controller.go

@@ -0,0 +1,76 @@
+/*
+Copyright 2018 The Kubernetes Authors.
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+*/
+
+package mci
+
+import (
+	"time"
+
+	"github.com/golang/glog"
+	"k8s.io/apimachinery/pkg/util/wait"
+	"k8s.io/client-go/tools/cache"
+	crv1alpha1 "k8s.io/cluster-registry/pkg/apis/clusterregistry/v1alpha1"
+	"k8s.io/ingress-gce/pkg/context"
+)
+
+// Controller is a barebones multi-cluster ingress controller.
+// For now, this controller only logs messages when CRUD operations are
+// made on a crv1alpha1.Cluster.
+type Controller struct {
+	resyncPeriod time.Duration
+
+	clusterSynced cache.InformerSynced
+	clusterLister cache.Indexer
+
+	// TODO(rramkumar): Add lister for service extension CRD.
+}
+
+func NewController(ctx *context.ControllerContext, resyncPeriod time.Duration) (*Controller, error) {
+	mciController := &Controller{
+		resyncPeriod:  resyncPeriod,
+		clusterSynced: ctx.MC.ClusterInformer.HasSynced,
+		clusterLister: ctx.MC.ClusterInformer.GetIndexer(),
+	}
+
+	ctx.MC.ClusterInformer.AddEventHandler(cache.ResourceEventHandlerFuncs{
+		AddFunc: func(obj interface{}) {
+			c := obj.(*crv1alpha1.Cluster)
+			glog.V(3).Infof("Cluster %v added", c.Name)
+		},
+		DeleteFunc: func(obj interface{}) {
+			c := obj.(*crv1alpha1.Cluster)
+			glog.V(3).Infof("Cluster %v deleted", c.Name)
+		},
+		UpdateFunc: func(obj, cur interface{}) {
+			c := obj.(*crv1alpha1.Cluster)
+			glog.V(3).Infof("Cluster %v updated", c.Name)
+		},
+	})
+	return mciController, nil
+}
+
+func (c *Controller) Run(stopCh <-chan struct{}) {
+	wait.PollUntil(5*time.Second, func() (bool, error) {
+		glog.V(2).Infof("Waiting for initial sync")
+		return c.synced(), nil
+	}, stopCh)
+
+	glog.V(2).Infof("Starting Multi-Cluster Ingress controller")
+}
+
+func (c *Controller) synced() bool {
+	return c.clusterSynced()
+}

pkg/neg/controller.go

@@ -27,7 +27,6 @@ import (
 	extensions "k8s.io/api/extensions/v1beta1"
 	"k8s.io/apimachinery/pkg/util/sets"
 	"k8s.io/apimachinery/pkg/util/wait"
-	"k8s.io/client-go/kubernetes"
 	"k8s.io/client-go/kubernetes/scheme"
 	unversionedcore "k8s.io/client-go/kubernetes/typed/core/v1"
 	"k8s.io/client-go/tools/cache"
@@ -62,7 +61,6 @@ type Controller struct {
 
 // NewController returns a network endpoint group controller.
 func NewController(
-	kubeClient kubernetes.Interface,
 	cloud networkEndpointGroupCloud,
 	ctx *context.ControllerContext,
 	zoneGetter zoneGetter,
@@ -74,7 +72,7 @@ func NewController(
 	eventBroadcaster := record.NewBroadcaster()
 	eventBroadcaster.StartLogging(glog.Infof)
 	eventBroadcaster.StartRecordingToSink(&unversionedcore.EventSinkImpl{
-		Interface: kubeClient.Core().Events(""),
+		Interface: ctx.KubeClient.Core().Events(""),
 	})
 	recorder := eventBroadcaster.NewRecorder(scheme.Scheme,
 		apiv1.EventSource{Component: "neg-controller"})

pkg/neg/controller_test.go

@@ -33,8 +33,8 @@ import (
 )
 
 func newTestController(kubeClient kubernetes.Interface) *Controller {
-	context := context.NewControllerContext(kubeClient, apiv1.NamespaceAll, 1*time.Second, true)
-	controller, _ := NewController(kubeClient,
+	context := context.NewControllerContext(kubeClient, nil, apiv1.NamespaceAll, 1*time.Second, true)
+	controller, _ := NewController(
 		NewFakeNetworkEndpointGroupCloud("test-subnetwork", "test-network"),
 		context,
 		NewFakeZoneGetter(),

pkg/neg/manager_test.go

@@ -36,7 +36,7 @@ const (
 )
 
 func NewTestSyncerManager(kubeClient kubernetes.Interface) *syncerManager {
-	context := context.NewControllerContext(kubeClient, apiv1.NamespaceAll, 1*time.Second, true)
+	context := context.NewControllerContext(kubeClient, nil, apiv1.NamespaceAll, 1*time.Second, true)
 	manager := newSyncerManager(
 		utils.NewNamer(CluseterID, ""),
 		record.NewFakeRecorder(100),

pkg/neg/syncer_test.go

@@ -23,7 +23,7 @@ const (
 
 func NewTestSyncer() *syncer {
 	kubeClient := fake.NewSimpleClientset()
-	context := context.NewControllerContext(kubeClient, apiv1.NamespaceAll, 1*time.Second, true)
+	context := context.NewControllerContext(kubeClient, nil, apiv1.NamespaceAll, 1*time.Second, true)
 	svcPort := servicePort{
 		namespace:  testServiceNamespace,
 		name:       testServiceName,