add provider spec for disabling default GKE provider behavior

config/crd/bases/uffizzi.com_uffizziclusters.yaml

@@ -154,6 +154,13 @@ spec:
                 type: object
               manifests:
                 type: string
+              provider:
+                default: vanila
+                enum:
+                - vanila
+                - gke
+                - eks
+                type: string
               resourceQuota:
                 description: UffizziClusterResourceQuota defines the resource quota
                   which defines the quota of resources a namespace has access to

main.go

@@ -63,7 +63,6 @@ func main() {
 		probeAddr                 string
 		enableLeaderElection      bool
 		concurrentReconciliations int
-		k8sProvider               string
 	)
 
 	flag.StringVar(&metricsAddr, "metrics-bind-address", ":8080", "The address the metric endpoint binds to.")
@@ -72,7 +71,6 @@ func main() {
 		"Enable leader election for controller manager. "+
 			"Enabling this will ensure there is only one active controller manager.")
 	flag.IntVar(&concurrentReconciliations, "concurrent", 5, "The number of concurrent reconciles per controller.")
-	flag.StringVar(&k8sProvider, "k8s-provider", "", "The k8s provider to use for the UffizziCluster")
 	opts := zap.Options{
 		Development: true,
 	}
@@ -101,9 +99,8 @@ func main() {
 	}
 	// Setup UffizziClusterReconciler
 	if err = (&uffizzicluster.UffizziClusterReconciler{
-		Client:             mgr.GetClient(),
-		Scheme:             mgr.GetScheme(),
-		KubernetesProvider: k8sProvider,
+		Client: mgr.GetClient(),
+		Scheme: mgr.GetScheme(),
 	}).SetupWithManager(mgr); err != nil {
 		setupLog.Error(err, "unable to create controller", "controller", "UffizziCluster")
 		os.Exit(1)

src/api/v1alpha1/uffizzicluster_types.go

@@ -143,7 +143,10 @@ type UffizziClusterResourceCount struct {
 type UffizziClusterSpec struct {
 	//+kubebuilder:default:="k3s"
 	//+kubebuilder:validation:Enum=k3s;k8s
-	Distro        string                       `json:"distro,omitempty"`
+	Distro string `json:"distro,omitempty"`
+	//+kubebuilder:default:="vanila"
+	//+kubebuilder:validation:Enum=vanila;gke;eks
+	Provider      string                       `json:"provider,omitempty"`
 	APIServer     UffizziClusterAPIServer      `json:"apiServer,omitempty"`
 	Ingress       UffizziClusterIngress        `json:"ingress,omitempty"`
 	TTL           string                       `json:"ttl,omitempty"`

src/controllers/uffizzicluster/uffizzicluster_controller.go

@@ -19,6 +19,7 @@
 import (
 	"context"
 	"encoding/json"
+	"fmt"
 	"github.com/UffizziCloud/uffizzi-cluster-operator/src/api/v1alpha1"
 	"github.com/UffizziCloud/uffizzi-cluster-operator/src/pkg/constants"
 	"github.com/UffizziCloud/uffizzi-cluster-operator/src/pkg/helm/build/vcluster"
@@ -42,8 +43,7 @@
 // UffizziClusterReconciler reconciles a UffizziCluster object
 type UffizziClusterReconciler struct {
 	client.Client
-	Scheme             *runtime.Scheme
-	KubernetesProvider string
+	Scheme *runtime.Scheme
 }
 
 //+kubebuilder:rbac:groups=uffizzi.com,resources=uffizziclusters,verbs=get;list;watch;create;update;patch;delete
@@ -250,15 +250,15 @@
 	// ----------------------
 	// UCLUSTER SLEEP
 	// ----------------------
-	//if err := r.reconcileSleepState(ctx, uCluster); err != nil {
-	//	if k8serrors.IsNotFound(err) {
-	//		// logger.Info("vcluster statefulset not found, requeueing")
-	//		return ctrl.Result{Requeue: true, RequeueAfter: time.Second * 5}, nil
-	//	}
-	//	// cluster did not sleep
-	//	logger.Info("Failed to reconcile sleep state, reconciling again", "Error", err.Error())
-	//	return ctrl.Result{Requeue: true, RequeueAfter: time.Second * 5}, nil
-	//}
+	if err := r.reconcileSleepState(ctx, uCluster); err != nil {
+		if k8serrors.IsNotFound(err) {
+			logger.Info("vcluster statefulset not found, will check again in the next round")
+			return ctrl.Result{}, nil
+		}
+		// cluster did not sleep
+		logger.Info("Failed to reconcile sleep state, reconciling again", "Error", err.Error())
+		return ctrl.Result{}, nil
+	}
 
 	return ctrl.Result{}, nil
 }
@@ -299,128 +299,87 @@
 // reconcileSleepState reconciles the sleep state of the vcluster
 // it also makes sure that the vcluster is up and running before setting the sleep state
 func (r *UffizziClusterReconciler) reconcileSleepState(ctx context.Context, uCluster *v1alpha1.UffizziCluster) error {
-	// get the patch copy of the uCluster so that we can have a good diff between the uCluster and patch object
-	//
-	patch := client.MergeFrom(uCluster.DeepCopy())
+	var (
+		// get the patch copy of the uCluster so that we can have a good diff between the uCluster and patch object
+		patch           = client.MergeFrom(uCluster.DeepCopy())
+		ucWorkload      runtime.Object
+		etcdStatefulSet *appsv1.StatefulSet
+		err             error
+	)
 	// get the stateful set or deployment created by the helm chart
-	ucWorkload, err := r.getUffizziClusterWorkload(ctx, uCluster)
-	if err != nil {
-		return err
+	// k3s is a statefulset, k8s is a deployment
+	if ucWorkload, err = r.getUffizziClusterWorkload(ctx, uCluster); err != nil {
+		return fmt.Errorf("failed to get uffizzicluster workload: %w", err)
 	}
 	// get the etcd stateful set created by the helm chart
-	etcdStatefulSet, err := r.getEtcdStatefulSet(ctx, uCluster)
+	if uCluster.Spec.ExternalDatastore == constants.ETCD {
+		if etcdStatefulSet, err = r.getEtcdStatefulSet(ctx, uCluster); err != nil {
+			return fmt.Errorf("failed to get etcd statefulset: %w", err)
+		}
+	}
 	// execute sleep reconciliation based on the type of workload
-	// TODO: Abstract the actual sleep reconciliation logic into a separate function so that it can be reused
-	// for different types of workloads, i.e. statefulset, deployment, daemonset
-	switch ucWorkload.(type) {
-	case *appsv1.StatefulSet:
-		ucStatefulSet := ucWorkload.(*appsv1.StatefulSet)
-		currentReplicas := ucStatefulSet.Spec.Replicas
-		// scale the vcluster instance to 0 if the sleep flag is true
-		if uCluster.Spec.Sleep && *currentReplicas > 0 {
-			var err error
-			if err = r.waitForStatefulSetToScale(ctx, 0, ucStatefulSet); err == nil {
-				setCondition(uCluster, APINotReady())
-			}
-			if uCluster.Spec.ExternalDatastore == constants.ETCD {
-				if err = r.waitForStatefulSetToScale(ctx, 0, etcdStatefulSet); err == nil {
-					setCondition(uCluster, DataStoreNotReady())
-				}
-			} else {
-				setCondition(uCluster, DataStoreNotReady())
-			}
-			if err != nil {
-				return err
-			}
-			err = r.deleteWorkloads(ctx, uCluster)
-			if err != nil {
-				return err
-			}
-			sleepingTime := metav1.Now().Rfc3339Copy()
-			setCondition(uCluster, Sleeping(sleepingTime))
-			// if the current replicas is 0, then do nothing
-		} else if !uCluster.Spec.Sleep && *currentReplicas == 0 {
-			statefulSets := []*appsv1.StatefulSet{}
-			statefulSets = append(statefulSets, ucStatefulSet)
-			if uCluster.Spec.ExternalDatastore == constants.ETCD {
-				statefulSets = append(statefulSets, etcdStatefulSet)
-			}
-			if err := r.scaleStatefulSets(ctx, 1, statefulSets...); err != nil {
-				return err
-			}
+	currentReplicas := r.getReplicasForWorkload(ucWorkload)
+	// scale the vcluster instance to 0 if the sleep flag is true
+	if uCluster.Spec.Sleep && currentReplicas > 0 {
+		if err = r.scaleWorkloads(ctx, 0, ucWorkload); err != nil {
+			return err
 		}
-		// ensure that the statefulset is up if the cluster is not sleeping
-		if !uCluster.Spec.Sleep {
-			// set status for vcluster waking up
-			lastAwakeTime := metav1.Now().Rfc3339Copy()
-			uCluster.Status.LastAwakeTime = lastAwakeTime
-			// if the above runs successfully, then set the status to awake
-			setCondition(uCluster, Awoken(lastAwakeTime))
-			var err error
-			if uCluster.Spec.ExternalDatastore == constants.ETCD {
-				if err = r.waitForStatefulSetToScale(ctx, 1, etcdStatefulSet); err == nil {
-					setCondition(uCluster, DataStoreReady())
-				}
-			} else {
-				setCondition(uCluster, DataStoreReady())
-			}
-			if err = r.waitForStatefulSetToScale(ctx, 1, ucStatefulSet); err == nil {
-				setCondition(uCluster, APIReady())
-			}
-			if err != nil {
-				return err
-			}
+		if err = r.waitForWorkloadToScale(ctx, 0, ucWorkload); err == nil {
+			setCondition(uCluster, APINotReady())
 		}
-	case *appsv1.Deployment:
-		ucDeployment := ucWorkload.(*appsv1.Deployment)
-		currentReplicas := ucDeployment.Spec.Replicas
-		// scale the vcluster instance to 0 if the sleep flag is true
-		if uCluster.Spec.Sleep && *currentReplicas > 0 {
-			var err error
-			if err = r.waitForDeploymentToScale(ctx, 0, ucDeployment); err == nil {
-				setCondition(uCluster, APINotReady())
-			}
-			if err = r.waitForStatefulSetToScale(ctx, 0, etcdStatefulSet); err == nil {
-				setCondition(uCluster, DataStoreNotReady())
-			}
-			if err != nil {
-				return err
-			}
-
-			err = r.deleteWorkloads(ctx, uCluster)
-			if err != nil {
+		if uCluster.Spec.ExternalDatastore == constants.ETCD {
+			if err = r.scaleWorkloads(ctx, 0, etcdStatefulSet); err != nil {
 				return err
 			}
-			sleepingTime := metav1.Now().Rfc3339Copy()
-			setCondition(uCluster, Sleeping(sleepingTime))
-			// if the current replicas is 0, then do nothing
-		} else if !uCluster.Spec.Sleep && *currentReplicas == 0 {
-			if err := r.scaleDeployments(ctx, 1, ucDeployment); err != nil {
-				return err
+			if err = r.waitForWorkloadToScale(ctx, 0, etcdStatefulSet); err == nil {
+				setCondition(uCluster, DataStoreNotReady())
 			}
+		} else {
+			setCondition(uCluster, DataStoreNotReady())
 		}
-		// ensure that the deployment is up if the cluster is not sleeping
-		if !uCluster.Spec.Sleep {
-			// set status for vcluster waking up
-			lastAwakeTime := metav1.Now().Rfc3339Copy()
-			uCluster.Status.LastAwakeTime = lastAwakeTime
-			// if the above runs successfully, then set the status to awake
-			setCondition(uCluster, Awoken(lastAwakeTime))
-			var err error
-			if err = r.waitForStatefulSetToScale(ctx, 1, etcdStatefulSet); err == nil {
-				setCondition(uCluster, APIReady())
-			}
-			if err = r.waitForDeploymentToScale(ctx, 1, ucDeployment); err == nil {
+		if err != nil {
+			return err
+		}
+		err = r.deleteWorkloads(ctx, uCluster)
+		if err != nil {
+			return err
+		}
+		sleepingTime := metav1.Now().Rfc3339Copy()
+		setCondition(uCluster, Sleeping(sleepingTime))
+		// if the current replicas is 0, then do nothing
+	} else if !uCluster.Spec.Sleep && currentReplicas == 0 {
+		workloads := []runtime.Object{}
+		workloads = append(workloads, ucWorkload)
+		if uCluster.Spec.ExternalDatastore == constants.ETCD {
+			workloads = append(workloads, etcdStatefulSet)
+		}
+		if err := r.scaleWorkloads(ctx, 1, workloads...); err != nil {
+			return err
+		}
+	}
+	// ensure that the statefulset is up if the cluster is not sleeping
+	if !uCluster.Spec.Sleep {
+		// set status for vcluster waking up
+		lastAwakeTime := metav1.Now().Rfc3339Copy()
+		uCluster.Status.LastAwakeTime = lastAwakeTime
+		// if the above runs successfully, then set the status to awake
+		setCondition(uCluster, Awoken(lastAwakeTime))
+		var err error
+		if uCluster.Spec.ExternalDatastore == constants.ETCD {
+			if err = r.waitForWorkloadToScale(ctx, 1, etcdStatefulSet); err == nil {
 				setCondition(uCluster, DataStoreReady())
 			}
-			if err != nil {
-				return err
-			}
+		} else {
+			setCondition(uCluster, DataStoreReady())
+		}
+		if err = r.waitForWorkloadToScale(ctx, 1, ucWorkload); err == nil {
+			setCondition(uCluster, APIReady())
+		}
+		if err != nil {
+			return err
 		}
-
-	default:
-		return errors.New("unknown workload type for vcluster")
 	}
+
 	if err := r.Status().Patch(ctx, uCluster, patch); err != nil {
 		return err
 	}

src/controllers/uffizzicluster/workload.go

@@ -2,9 +2,11 @@
 
 import (
 	context "context"
+	"fmt"
 	"github.com/UffizziCloud/uffizzi-cluster-operator/src/api/v1alpha1"
 	"github.com/UffizziCloud/uffizzi-cluster-operator/src/pkg/constants"
 	"github.com/UffizziCloud/uffizzi-cluster-operator/src/pkg/helm/build/vcluster"
+	appsv1 "k8s.io/api/apps/v1"
 	corev1 "k8s.io/api/core/v1"
 	"k8s.io/apimachinery/pkg/runtime"
 	"sigs.k8s.io/controller-runtime/pkg/client"
@@ -33,3 +35,63 @@
 	}
 	return nil
 }
+
+func (r *UffizziClusterReconciler) scaleWorkloads(ctx context.Context, scale int, workloads ...runtime.Object) error {
+	sss := []*appsv1.StatefulSet{}
+	ds := []*appsv1.Deployment{}
+	for _, w := range workloads {
+		if w != nil {
+			switch w.(type) {
+			case *appsv1.StatefulSet:
+				ss := w.(*appsv1.StatefulSet)
+				sss = append(sss, ss)
+			case *appsv1.Deployment:
+				d := w.(*appsv1.Deployment)
+				ds = append(ds, d)
+			}
+		}
+	}
+	if len(sss) > 0 {
+		if err := r.scaleStatefulSets(ctx, scale, sss...); err != nil {
+			return fmt.Errorf("failed to scale stateful sets: %w", err)
+		}
+	}
+	if len(ds) > 0 {
+		if err := r.scaleDeployments(ctx, scale, ds...); err != nil {
+			return fmt.Errorf("failed to scale deployments: %w", err)
+		}
+	}
+	return nil
+}
+
+func (r *UffizziClusterReconciler) waitForWorkloadToScale(ctx context.Context, scale int, w runtime.Object) error {
+	if w != nil {
+		switch w.(type) {
+		case *appsv1.StatefulSet:
+			ss := w.(*appsv1.StatefulSet)
+			if err := r.waitForStatefulSetToScale(ctx, scale, ss); err != nil {
+				return fmt.Errorf("failed to wait for stateful sets to scale: %w", err)
+			}
+		case *appsv1.Deployment:
+			d := w.(*appsv1.Deployment)
+			if err := r.waitForDeploymentToScale(ctx, scale, d); err != nil {
+				return fmt.Errorf("failed to wait for deployments to scale: %w", err)
+			}
+		}
+	}
+	return nil
+}
+
+func (r *UffizziClusterReconciler) getReplicasForWorkload(w runtime.Object) int {
+	if w != nil {
+		switch w.(type) {
+		case *appsv1.StatefulSet:
+			ss := w.(*appsv1.StatefulSet)
+			return int(*ss.Spec.Replicas)
+		case *appsv1.Deployment:
+			d := w.(*appsv1.Deployment)
+			return int(*d.Spec.Replicas)
+		}
+	}
+	return 0
+}

src/pkg/constants/constants.go

@@ -17,6 +17,9 @@ const (
 	LOFT_CHART_REPO_URL         = "https://charts.loft.sh"
 	VCLUSTER_K3S_DISTRO         = "k3s"
 	VCLUSTER_K8S_DISTRO         = "k8s"
+	PROVIDER_NOTHING            = "vanila"
+	PROVIDER_GKE                = "gke"
+	PROVIDER_EKS                = "eks"
 	K3S_DATASTORE_ENDPOINT      = "K3S_DATASTORE_ENDPOINT"
 	VCLUSTER_INGRESS_HOSTNAME   = "VCLUSTER_INGRESS_HOST"
 	DEFAULT_K3S_VERSION         = "rancher/k3s:v1.27.3-k3s1"
@@ -31,13 +34,9 @@ const (
 )
 
 type LIFECYCLE_OP_TYPE string
-type KUBERNETES_PROVIDER string
 
 const (
-	LIFECYCLE_OP_TYPE_CREATE LIFECYCLE_OP_TYPE   = "create"
-	LIFECYCLE_OP_TYPE_UPDATE LIFECYCLE_OP_TYPE   = "update"
-	LIFECYCLE_OP_TYPE_DELETE LIFECYCLE_OP_TYPE   = "delete"
-	NO_KUBE_PROVIDER         KUBERNETES_PROVIDER = ""
-	GKE_KUBE_PROVIDER        KUBERNETES_PROVIDER = "gke"
-	EKS_KUBE_PROVIDER        KUBERNETES_PROVIDER = "eks"
+	LIFECYCLE_OP_TYPE_CREATE LIFECYCLE_OP_TYPE = "create"
+	LIFECYCLE_OP_TYPE_UPDATE LIFECYCLE_OP_TYPE = "update"
+	LIFECYCLE_OP_TYPE_DELETE LIFECYCLE_OP_TYPE = "delete"
 )