health: add liveness and readiness probes (#508)

Author: cpanato

config/manager/manager.yaml

@@ -28,6 +28,18 @@ spec:
         image: controller:latest
         imagePullPolicy: Always
         name: manager
+        ports:
+        - containerPort: 9440
+          name: healthz
+          protocol: TCP
+        readinessProbe:
+          httpGet:
+            path: /readyz
+            port: healthz
+        livenessProbe:
+          httpGet:
+            path: /healthz
+            port: healthz
       terminationGracePeriodSeconds: 10
       tolerations:
         - effect: NoSchedule

main.go

@@ -26,6 +26,7 @@ import (
 
 	"sigs.k8s.io/cluster-api-provider-openstack/pkg/record"
 	"sigs.k8s.io/controller-runtime/pkg/client/config"
+	"sigs.k8s.io/controller-runtime/pkg/healthz"
 
 	_ "net/http/pprof"
 
@@ -63,6 +64,7 @@ func main() {
 		openStackClusterConcurrency int
 		openStackMachineConcurrency int
 		syncPeriod                  time.Duration
+		healthAddr                  string
 	)
 
 	flag.StringVar(
@@ -111,6 +113,12 @@ func main() {
 		"The minimum interval at which watched resources are reconciled (e.g. 15m)",
 	)
 
+	flag.StringVar(&healthAddr,
+		"health-addr",
+		":9440",
+		"The address the health endpoint binds to.",
+	)
+
 	flag.Parse()
 
 	if watchNamespace != "" {
@@ -132,12 +140,13 @@ func main() {
 	}
 
 	mgr, err := ctrl.NewManager(cfg, ctrl.Options{
-		Scheme:             scheme,
-		MetricsBindAddress: metricsAddr,
-		LeaderElection:     enableLeaderElection,
-		LeaderElectionID:   "controller-leader-election-capo",
-		SyncPeriod:         &syncPeriod,
-		Namespace:          watchNamespace,
+		Scheme:                 scheme,
+		MetricsBindAddress:     metricsAddr,
+		LeaderElection:         enableLeaderElection,
+		LeaderElectionID:       "controller-leader-election-capo",
+		SyncPeriod:             &syncPeriod,
+		Namespace:              watchNamespace,
+		HealthProbeBindAddress: healthAddr,
 	})
 	if err != nil {
 		setupLog.Error(err, "unable to start manager")
@@ -165,6 +174,16 @@ func main() {
 	}
 	// +kubebuilder:scaffold:builder
 
+	if err := mgr.AddReadyzCheck("ping", healthz.Ping); err != nil {
+		setupLog.Error(err, "unable to create ready check")
+		os.Exit(1)
+	}
+
+	if err := mgr.AddHealthzCheck("ping", healthz.Ping); err != nil {
+		setupLog.Error(err, "unable to create health check")
+		os.Exit(1)
+	}
+
 	setupLog.Info("starting manager")
 	if err := mgr.Start(ctrl.SetupSignalHandler()); err != nil {
 		setupLog.Error(err, "problem running manager")