Operator developers can take advantage of Go programming language support in the Operator SDK to build an example Go-based Operator for Memcached, a distributed key-value store, and manage its lifecycle.

This process is accomplished using two centerpieces of the Operator Framework:

Operator SDK

The operator-sdk CLI tool and controller-runtime library API

Operator Lifecycle Manager (OLM)

Installation, upgrade, and role-based access control (RBAC) of Operators on a cluster

This tutorial goes into greater detail than Getting started with Operator SDK for Go-based Operators.


  • Operator SDK CLI installed

  • OpenShift CLI (oc) v4.10+ installed

  • Logged into an OpenShift Container Platform 4.10 cluster with oc with an account that has cluster-admin permissions

  • To allow the cluster to pull the image, the repository where you push your image must be set as public, or you must configure an image pull secret

Creating a project

Use the Operator SDK CLI to create a project called memcached-operator.

  1. Create a directory for the project:

    $ mkdir -p $HOME/projects/memcached-operator
  2. Change to the directory:

    $ cd $HOME/projects/memcached-operator
  3. Activate support for Go modules:

    $ export GO111MODULE=on
  4. Run the operator-sdk init command to initialize the project:

    $ operator-sdk init \
        --domain=example.com \

    The operator-sdk init command uses the Go plug-in by default.

    The operator-sdk init command generates a go.mod file to be used with Go modules. The --repo flag is required when creating a project outside of $GOPATH/src/, because generated files require a valid module path.


Among the files generated by the operator-sdk init command is a Kubebuilder PROJECT file. Subsequent operator-sdk commands, as well as help output, that are run from the project root read this file and are aware that the project type is Go. For example:

domain: example.com
layout: go.kubebuilder.io/v3
projectName: memcached-operator
repo: github.com/example-inc/memcached-operator
version: 3
  manifests.sdk.operatorframework.io/v2: {}
  scorecard.sdk.operatorframework.io/v2: {}

About the Manager

The main program for the Operator is the main.go file, which initializes and runs the Manager. The Manager automatically registers the Scheme for all custom resource (CR) API definitions and sets up and runs controllers and webhooks.

The Manager can restrict the namespace that all controllers watch for resources:

mgr, err := ctrl.NewManager(cfg, manager.Options{Namespace: namespace})

By default, the Manager watches the namespace where the Operator runs. To watch all namespaces, you can leave the namespace option empty:

mgr, err := ctrl.NewManager(cfg, manager.Options{Namespace: ""})

You can also use the MultiNamespacedCacheBuilder function to watch a specific set of namespaces:

var namespaces []string (1)
mgr, err := ctrl.NewManager(cfg, manager.Options{ (2)
   NewCache: cache.MultiNamespacedCacheBuilder(namespaces),
1 List of namespaces.
2 Creates a Cmd struct to provide shared dependencies and start components.

About multi-group APIs

Before you create an API and controller, consider whether your Operator requires multiple API groups. This tutorial covers the default case of a single group API, but to change the layout of your project to support multi-group APIs, you can run the following command:

$ operator-sdk edit --multigroup=true

This command updates the PROJECT file, which should look like the following example:

domain: example.com
layout: go.kubebuilder.io/v3
multigroup: true

For multi-group projects, the API Go type files are created in the apis/<group>/<version>/ directory, and the controllers are created in the controllers/<group>/ directory. The Dockerfile is then updated accordingly.

Additional resource

Creating an API and controller

Use the Operator SDK CLI to create a custom resource definition (CRD) API and controller.

  1. Run the following command to create an API with group cache, version, v1, and kind Memcached:

    $ operator-sdk create api \
        --group=cache \
        --version=v1 \
  2. When prompted, enter y for creating both the resource and controller:

    Create Resource [y/n]
    Create Controller [y/n]
    Example output
    Writing scaffold for you to edit...

This process generates the Memcached resource API at api/v1/memcached_types.go and the controller at controllers/memcached_controller.go.

Defining the API

Define the API for the Memcached custom resource (CR).

  1. Modify the Go type definitions at api/v1/memcached_types.go to have the following spec and status:

    // MemcachedSpec defines the desired state of Memcached
    type MemcachedSpec struct {
    	// +kubebuilder:validation:Minimum=0
    	// Size is the size of the memcached deployment
    	Size int32 `json:"size"`
    // MemcachedStatus defines the observed state of Memcached
    type MemcachedStatus struct {
    	// Nodes are the names of the memcached pods
    	Nodes []string `json:"nodes"`
  2. Update the generated code for the resource type:

    $ make generate

    After you modify a *_types.go file, you must run the make generate command to update the generated code for that resource type.

    The above Makefile target invokes the controller-gen utility to update the api/v1/zz_generated.deepcopy.go file. This ensures your API Go type definitions implement the runtime.Object interface that all Kind types must implement.

Generating CRD manifests

After the API is defined with spec and status fields and custom resource definition (CRD) validation markers, you can generate CRD manifests.

  • Run the following command to generate and update CRD manifests:

    $ make manifests

    This Makefile target invokes the controller-gen utility to generate the CRD manifests in the config/crd/bases/cache.example.com_memcacheds.yaml file.

About OpenAPI validation

OpenAPIv3 schemas are added to CRD manifests in the spec.validation block when the manifests are generated. This validation block allows Kubernetes to validate the properties in a Memcached custom resource (CR) when it is created or updated.

Markers, or annotations, are available to configure validations for your API. These markers always have a +kubebuilder:validation prefix.

Additional resources

Implementing the controller

After creating a new API and controller, you can implement the controller logic.

  • For this example, replace the generated controller file controllers/memcached_controller.go with following example implementation:

    Example memcached_controller.go
    Copyright 2020.
    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
    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 controllers
    import (
    	appsv1 "k8s.io/api/apps/v1"
    	corev1 "k8s.io/api/core/v1"
    	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
    	ctrl "sigs.k8s.io/controller-runtime"
    	cachev1alpha1 "github.com/example/memcached-operator/api/v1alpha1"
    // MemcachedReconciler reconciles a Memcached object
    type MemcachedReconciler struct {
    	Log    logr.Logger
    	Scheme *runtime.Scheme
    // +kubebuilder:rbac:groups=cache.example.com,resources=memcacheds,verbs=get;list;watch;create;update;patch;delete
    // +kubebuilder:rbac:groups=cache.example.com,resources=memcacheds/status,verbs=get;update;patch
    // +kubebuilder:rbac:groups=cache.example.com,resources=memcacheds/finalizers,verbs=update
    // +kubebuilder:rbac:groups=apps,resources=deployments,verbs=get;list;watch;create;update;patch;delete
    // +kubebuilder:rbac:groups=core,resources=pods,verbs=get;list;
    // Reconcile is part of the main kubernetes reconciliation loop which aims to
    // move the current state of the cluster closer to the desired state.
    // TODO(user): Modify the Reconcile function to compare the state specified by
    // the Memcached object against the actual cluster state, and then
    // perform operations to make the cluster state reflect the state specified by
    // the user.
    // For more details, check Reconcile and its Result here:
    // - https://pkg.go.dev/sigs.k8s.io/controller-runtime@v0.7.0/pkg/reconcile
    func (r *MemcachedReconciler) Reconcile(ctx context.Context, req ctrl.Request) (ctrl.Result, error) {
    	log := r.Log.WithValues("memcached", req.NamespacedName)
    	// Fetch the Memcached instance
    	memcached := &cachev1alpha1.Memcached{}
    	err := r.Get(ctx, req.NamespacedName, memcached)
    	if err != nil {
    		if errors.IsNotFound(err) {
    			// Request object not found, could have been deleted after reconcile request.
    			// Owned objects are automatically garbage collected. For additional cleanup logic use finalizers.
    			// Return and don't requeue
    			log.Info("Memcached resource not found. Ignoring since object must be deleted")
    			return ctrl.Result{}, nil
    		// Error reading the object - requeue the request.
    		log.Error(err, "Failed to get Memcached")
    		return ctrl.Result{}, err
    	// Check if the deployment already exists, if not create a new one
    	found := &appsv1.Deployment{}
    	err = r.Get(ctx, types.NamespacedName{Name: memcached.Name, Namespace: memcached.Namespace}, found)
    	if err != nil && errors.IsNotFound(err) {
    		// Define a new deployment
    		dep := r.deploymentForMemcached(memcached)
    		log.Info("Creating a new Deployment", "Deployment.Namespace", dep.Namespace, "Deployment.Name", dep.Name)
    		err = r.Create(ctx, dep)
    		if err != nil {
    			log.Error(err, "Failed to create new Deployment", "Deployment.Namespace", dep.Namespace, "Deployment.Name", dep.Name)
    			return ctrl.Result{}, err
    		// Deployment created successfully - return and requeue
    		return ctrl.Result{Requeue: true}, nil
    	} else if err != nil {
    		log.Error(err, "Failed to get Deployment")
    		return ctrl.Result{}, err
    	// Ensure the deployment size is the same as the spec
    	size := memcached.Spec.Size
    	if *found.Spec.Replicas != size {
    		found.Spec.Replicas = &size
    		err = r.Update(ctx, found)
    		if err != nil {
    			log.Error(err, "Failed to update Deployment", "Deployment.Namespace", found.Namespace, "Deployment.Name", found.Name)
    			return ctrl.Result{}, err
    		// Spec updated - return and requeue
    		return ctrl.Result{Requeue: true}, nil
    	// Update the Memcached status with the pod names
    	// List the pods for this memcached's deployment
    	podList := &corev1.PodList{}
    	listOpts := []client.ListOption{
    	if err = r.List(ctx, podList, listOpts...); err != nil {
    		log.Error(err, "Failed to list pods", "Memcached.Namespace", memcached.Namespace, "Memcached.Name", memcached.Name)
    		return ctrl.Result{}, err
    	podNames := getPodNames(podList.Items)
    	// Update status.Nodes if needed
    	if !reflect.DeepEqual(podNames, memcached.Status.Nodes) {
    		memcached.Status.Nodes = podNames
    		err := r.Status().Update(ctx, memcached)
    		if err != nil {
    			log.Error(err, "Failed to update Memcached status")
    			return ctrl.Result{}, err
    	return ctrl.Result{}, nil
    // deploymentForMemcached returns a memcached Deployment object
    func (r *MemcachedReconciler) deploymentForMemcached(m *cachev1alpha1.Memcached) *appsv1.Deployment {
    	ls := labelsForMemcached(m.Name)
    	replicas := m.Spec.Size
    	dep := &appsv1.Deployment{
    		ObjectMeta: metav1.ObjectMeta{
    			Name:      m.Name,
    			Namespace: m.Namespace,
    		Spec: appsv1.DeploymentSpec{
    			Replicas: &replicas,
    			Selector: &metav1.LabelSelector{
    				MatchLabels: ls,
    			Template: corev1.PodTemplateSpec{
    				ObjectMeta: metav1.ObjectMeta{
    					Labels: ls,
    				Spec: corev1.PodSpec{
    					Containers: []corev1.Container{{
    						Image:   "memcached:1.4.36-alpine",
    						Name:    "memcached",
    						Command: []string{"memcached", "-m=64", "-o", "modern", "-v"},
    						Ports: []corev1.ContainerPort{{
    							ContainerPort: 11211,
    							Name:          "memcached",
    	// Set Memcached instance as the owner and controller
    	ctrl.SetControllerReference(m, dep, r.Scheme)
    	return dep
    // labelsForMemcached returns the labels for selecting the resources
    // belonging to the given memcached CR name.
    func labelsForMemcached(name string) map[string]string {
    	return map[string]string{"app": "memcached", "memcached_cr": name}
    // getPodNames returns the pod names of the array of pods passed in
    func getPodNames(pods []corev1.Pod) []string {
    	var podNames []string
    	for _, pod := range pods {
    		podNames = append(podNames, pod.Name)
    	return podNames
    // SetupWithManager sets up the controller with the Manager.
    func (r *MemcachedReconciler) SetupWithManager(mgr ctrl.Manager) error {
    	return ctrl.NewControllerManagedBy(mgr).

    The example controller runs the following reconciliation logic for each Memcached custom resource (CR):

    • Create a Memcached deployment if it does not exist.

    • Ensure that the deployment size is the same as specified by the Memcached CR spec.

    • Update the Memcached CR status with the names of the memcached pods.

The next subsections explain how the controller in the example implementation watches resources and how the reconcile loop is triggered. You can skip these subsections to go directly to Running the Operator.

Resources watched by the controller

The SetupWithManager() function in controllers/memcached_controller.go specifies how the controller is built to watch a CR and other resources that are owned and managed by that controller.

import (
	appsv1 "k8s.io/api/apps/v1"

func (r *MemcachedReconciler) SetupWithManager(mgr ctrl.Manager) error {
	return ctrl.NewControllerManagedBy(mgr).

NewControllerManagedBy() provides a controller builder that allows various controller configurations.

For(&cachev1.Memcached{}) specifies the Memcached type as the primary resource to watch. For each Add, Update, or Delete event for a Memcached type, the reconcile loop is sent a reconcile Request argument, which consists of a namespace and name key, for that Memcached object.

Owns(&appsv1.Deployment{}) specifies the Deployment type as the secondary resource to watch. For each Deployment type Add, Update, or Delete event, the event handler maps each event to a reconcile request for the owner of the deployment. In this case, the owner is the Memcached object for which the deployment was created.

Controller configurations

You can initialize a controller by using many other useful configurations. For example:

  • Set the maximum number of concurrent reconciles for the controller by using the MaxConcurrentReconciles option, which defaults to 1:

    func (r *MemcachedReconciler) SetupWithManager(mgr ctrl.Manager) error {
        return ctrl.NewControllerManagedBy(mgr).
                MaxConcurrentReconciles: 2,
  • Filter watch events using predicates.

  • Choose the type of EventHandler to change how a watch event translates to reconcile requests for the reconcile loop. For Operator relationships that are more complex than primary and secondary resources, you can use the EnqueueRequestsFromMapFunc handler to transform a watch event into an arbitrary set of reconcile requests.

For more details on these and other configurations, see the upstream Builder and Controller GoDocs.

Reconcile loop

Every controller has a reconciler object with a Reconcile() method that implements the reconcile loop. The reconcile loop is passed the Request argument, which is a namespace and name key used to find the primary resource object, Memcached, from the cache:

import (
	ctrl "sigs.k8s.io/controller-runtime"

	cachev1 "github.com/example-inc/memcached-operator/api/v1"

func (r *MemcachedReconciler) Reconcile(ctx context.Context, req ctrl.Request) (ctrl.Result, error) {
  // Lookup the Memcached instance for this reconcile request
  memcached := &cachev1.Memcached{}
  err := r.Get(ctx, req.NamespacedName, memcached)

Based on the return values, result, and error, the request might be requeued and the reconcile loop might be triggered again:

// Reconcile successful - don't requeue
return ctrl.Result{}, nil
// Reconcile failed due to error - requeue
return ctrl.Result{}, err
// Requeue for any reason other than an error
return ctrl.Result{Requeue: true}, nil

You can set the Result.RequeueAfter to requeue the request after a grace period as well:

import "time"

// Reconcile for any reason other than an error after 5 seconds
return ctrl.Result{RequeueAfter: time.Second*5}, nil

You can return Result with RequeueAfter set to periodically reconcile a CR.

For more on reconcilers, clients, and interacting with resource events, see the Controller Runtime Client API documentation.

Permissions and RBAC manifests

The controller requires certain RBAC permissions to interact with the resources it manages. These are specified using RBAC markers, such as the following:

// +kubebuilder:rbac:groups=cache.example.com,resources=memcacheds,verbs=get;list;watch;create;update;patch;delete
// +kubebuilder:rbac:groups=cache.example.com,resources=memcacheds/status,verbs=get;update;patch
// +kubebuilder:rbac:groups=cache.example.com,resources=memcacheds/finalizers,verbs=update
// +kubebuilder:rbac:groups=apps,resources=deployments,verbs=get;list;watch;create;update;patch;delete
// +kubebuilder:rbac:groups=core,resources=pods,verbs=get;list;

func (r *MemcachedReconciler) Reconcile(ctx context.Context, req ctrl.Request) (ctrl.Result, error) {

The ClusterRole object manifest at config/rbac/role.yaml is generated from the previous markers by using the controller-gen utility whenever the make manifests command is run.

Enabling proxy support

Operator authors can develop Operators that support network proxies. Cluster administrators configure proxy support for the environment variables that are handled by Operator Lifecycle Manager (OLM). To support proxied clusters, your Operator must inspect the environment for the following standard proxy variables and pass the values to Operands:




This tutorial uses HTTP_PROXY as an example environment variable.

  • A cluster with cluster-wide egress proxy enabled.