KUBESAW-12: Convert the health-check goroutine into ToolchainCluster controller

controllers/toolchainclustercache/healthchecker.go → controllers/toolchainclusterhealth/healthchecker.go

@@ -1,92 +1,32 @@
-package toolchainclustercache
+package toolchainclusterhealth
 
 import (
 	"context"
-	"fmt"
 	"strings"
-	"time"
 
 	toolchainv1alpha1 "github.com/codeready-toolchain/api/api/v1alpha1"
-	"github.com/codeready-toolchain/toolchain-common/pkg/cluster"
 	"github.com/go-logr/logr"
 	"github.com/pkg/errors"
 	corev1 "k8s.io/api/core/v1"
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
-	"k8s.io/apimachinery/pkg/util/wait"
 	kubeclientset "k8s.io/client-go/kubernetes"
 	"sigs.k8s.io/controller-runtime/pkg/client"
-	logf "sigs.k8s.io/controller-runtime/pkg/log"
-	"sigs.k8s.io/controller-runtime/pkg/manager"
 )
 
-var logger = logf.Log.WithName("toolchaincluster_healthcheck")
-
 const (
 	healthzOk              = "/healthz responded with ok"
 	healthzNotOk           = "/healthz responded without ok"
 	clusterNotReachableMsg = "cluster is not reachable"
 	clusterReachableMsg    = "cluster is reachable"
 )
 
-func StartHealthChecks(ctx context.Context, mgr manager.Manager, namespace string, period time.Duration) {
-	logger.Info("starting health checks", "period", period)
-	go wait.Until(func() {
-		updateClusterStatuses(ctx, namespace, mgr.GetClient())
-	}, period, ctx.Done())
-}
-
 type HealthChecker struct {
 	localClusterClient     client.Client
 	remoteClusterClient    client.Client
 	remoteClusterClientset *kubeclientset.Clientset
 	logger                 logr.Logger
 }
 
-// updateClusterStatuses checks cluster health and updates status of all ToolchainClusters
-func updateClusterStatuses(ctx context.Context, namespace string, cl client.Client) {
-	clusters := &toolchainv1alpha1.ToolchainClusterList{}
-	err := cl.List(ctx, clusters, client.InNamespace(namespace))
-	if err != nil {
-		logger.Error(err, "unable to list existing ToolchainClusters")
-		return
-	}
-	if len(clusters.Items) == 0 {
-		logger.Info("no ToolchainCluster found")
-	}
-
-	for _, obj := range clusters.Items {
-		clusterObj := obj.DeepCopy()
-		clusterLogger := logger.WithValues("cluster-name", clusterObj.Name)
-
-		cachedCluster, ok := cluster.GetCachedToolchainCluster(clusterObj.Name)
-		if !ok {
-			clusterLogger.Error(fmt.Errorf("cluster %s not found in cache", clusterObj.Name), "failed to retrieve stored data for cluster")
-			clusterObj.Status.Conditions = []toolchainv1alpha1.ToolchainClusterCondition{clusterOfflineCondition()}
-			if err := cl.Status().Update(ctx, clusterObj); err != nil {
-				clusterLogger.Error(err, "failed to update the status of ToolchainCluster")
-			}
-			continue
-		}
-
-		clientSet, err := kubeclientset.NewForConfig(cachedCluster.RestConfig)
-		if err != nil {
-			clusterLogger.Error(err, "cannot create ClientSet for a ToolchainCluster")
-			continue
-		}
-
-		healthChecker := &HealthChecker{
-			localClusterClient:     cl,
-			remoteClusterClient:    cachedCluster.Client,
-			remoteClusterClientset: clientSet,
-			logger:                 clusterLogger,
-		}
-		// clusterLogger.Info("getting the current state of ToolchainCluster")
-		if err := healthChecker.updateIndividualClusterStatus(ctx, clusterObj); err != nil {
-			clusterLogger.Error(err, "unable to update cluster status of ToolchainCluster")
-		}
-	}
-}
-
 func (hc *HealthChecker) updateIndividualClusterStatus(ctx context.Context, toolchainCluster *toolchainv1alpha1.ToolchainCluster) error {
 
 	currentClusterStatus := hc.getClusterHealthStatus(ctx)

controllers/toolchainclusterhealth/toolchaincluster_healthcheck_controller.go

@@ -0,0 +1,97 @@
+package toolchainclusterhealth
+
+import (
+	"context"
+	"fmt"
+	"time"
+
+	toolchainv1alpha1 "github.com/codeready-toolchain/api/api/v1alpha1"
+	"github.com/codeready-toolchain/toolchain-common/pkg/cluster"
+	"k8s.io/apimachinery/pkg/api/errors"
+	"k8s.io/apimachinery/pkg/runtime"
+	kubeclientset "k8s.io/client-go/kubernetes"
+	ctrl "sigs.k8s.io/controller-runtime"
+	"sigs.k8s.io/controller-runtime/pkg/client"
+	"sigs.k8s.io/controller-runtime/pkg/log"
+	"sigs.k8s.io/controller-runtime/pkg/manager"
+	"sigs.k8s.io/controller-runtime/pkg/reconcile"
+)
+
+// NewReconciler returns a new Reconciler
+func NewReconciler(mgr manager.Manager, namespace string, timeout time.Duration, requeAfter time.Duration) *Reconciler {
+	cacheLog := log.Log.WithName("toolchaincluster_health")
+	clusterCacheService := cluster.NewToolchainClusterService(mgr.GetClient(), cacheLog, namespace, timeout)
+	return &Reconciler{
+		client:              mgr.GetClient(),
+		scheme:              mgr.GetScheme(),
+		clusterCacheService: clusterCacheService,
+		requeAfter:          requeAfter,
+	}
+}
+
+// SetupWithManager sets up the controller with the Manager.
+func (r *Reconciler) SetupWithManager(mgr ctrl.Manager) error {
+	return ctrl.NewControllerManagedBy(mgr).
+		For(&toolchainv1alpha1.ToolchainCluster{}).
+		Complete(r)
+}
+
+// Reconciler reconciles a ToolchainCluster object
+type Reconciler struct {
+	client              client.Client
+	scheme              *runtime.Scheme
+	clusterCacheService cluster.ToolchainClusterService
+	requeAfter          time.Duration
+}
+
+// Reconcile reads that state of the cluster for a ToolchainCluster object and makes changes based on the state read
+// and what is in the ToolchainCluster.Spec. It updates the status of the individual cluster
+// Note:
+// The Controller will requeue the Request to be processed again if the returned error is non-nil or
+// Result.Requeue is true, otherwise upon completion it will remove the work from the queue.
+func (r *Reconciler) Reconcile(ctx context.Context, request ctrl.Request) (ctrl.Result, error) {
+	reqLogger := log.FromContext(ctx).WithName("health")
+	reqLogger.Info("Reconciling ToolchainCluster")
+
+	// Fetch the ToolchainCluster instance
+	toolchainCluster := &toolchainv1alpha1.ToolchainCluster{}
+	err := r.client.Get(ctx, request.NamespacedName, toolchainCluster)
+	if err != nil {
+		if errors.IsNotFound(err) {
+			// Stop monitoring the toolchain cluster as it is deleted
+			return reconcile.Result{}, nil
+		}
+		// Error reading the object - requeue the request.
+		return reconcile.Result{}, err
+	}
+
+	cachedCluster, ok := cluster.GetCachedToolchainCluster(toolchainCluster.Name)
+	if !ok {
+		err := fmt.Errorf("cluster %s not found in cache", toolchainCluster.Name)
+		toolchainCluster.Status.Conditions = []toolchainv1alpha1.ToolchainClusterCondition{clusterOfflineCondition()}
+		if err := r.client.Status().Update(ctx, toolchainCluster); err != nil {
+			reqLogger.Error(err, "failed to update the status of ToolchainCluster")
+		}
+		return reconcile.Result{}, err
+	}
+
+	clientSet, err := kubeclientset.NewForConfig(cachedCluster.RestConfig)
+	if err != nil {
+		return reconcile.Result{}, err
+	}
+
+	healthChecker := &HealthChecker{
+		localClusterClient:     r.client,
+		remoteClusterClient:    cachedCluster.Client,
+		remoteClusterClientset: clientSet,
+		logger:                 reqLogger,
+	}
+
+	//update the status of the individual cluster.
+	if err := healthChecker.updateIndividualClusterStatus(ctx, toolchainCluster); err != nil {
+		reqLogger.Error(err, "unable to update cluster status of ToolchainCluster")
+		return reconcile.Result{}, err
+	}
+
+	return reconcile.Result{RequeueAfter: r.requeAfter}, nil
+}