Need help with getting-started?
Click the “chat” button below for chat support from the developer who created it, or find similar developers for support.

About the developer

535 Stars 112 Forks 80 Commits 0 Opened issues


An end-to-end scenario showing how to use the Operator Framework.

Services available


Need anything else?

Contributors list

[DEPRECATED] Getting Started

This project is deprecated. Documentation for the Operator Framework has moved to


The Operator Framework (intro blog post) is an open source toolkit to manage Kubernetes native applications, called operators, in an effective, automated, and scalable way. Operators take advantage of Kubernetes's extensibility to deliver the automation advantages of cloud services like provisioning, scaling, and backup/restore while being able to run anywhere that Kubernetes can run.

This guide shows how to build a simple memcached operator and how to manage its lifecycle from install to update to a new version. For that, we will use two center pieces of the framework:

  • Operator SDK: Allows your developers to build an operator based on your expertise without requiring knowledge of Kubernetes API complexities.
  • Operator Lifecycle Manager: Helps you to install, update, and generally manage the lifecycle of all of the operators (and their associated services) running across your clusters.

Build an operator using the Operator SDK

BEFORE YOU BEGIN: links to the Operator SDK repo in this document are pinned to the

branch. Make sure you update the link such that it points to the correct Operator SDK repo version, which should match this repo's version or the
operator-sdk version
being used. For example, if you are using
v0.12.0, update all links from this repo to the SDK repo with
master -> v0.12.0
. Otherwise you may see incorrect information.

The Operator SDK makes it easier to build Kubernetes native applications, a process that can require deep, application-specific operational knowledge. The SDK not only lowers that barrier, but it also helps reduce the amount of boilerplate code needed for many common management capabilities, such as metering or monitoring.

This section walks through an example of building a simple memcached operator using tools and libraries provided by the Operator SDK. This walkthrough is not exhaustive; for an in-depth explanation of these steps, see the SDK's user guide.

Requirements: Please make sure that the Operator SDK is installed on the development machine. Additionally, the Operator Lifecycle Manager must be installed in the cluster (1.8 or above to support the apps/v1beta2 API group) before running this guide.

Create a new project

  1. Use the CLI to create a new
$ mkdir -p $GOPATH/src/
$ cd $GOPATH/src/
$ export GO111MODULE=on
$ operator-sdk new memcached-operator
$ cd memcached-operator

This creates the

  1. Install dependencies by running
    go mod tidy

NOTE: Learn more about the project directory structure from the SDK project layout documentation.


The main program for the operator

initializes and runs the Manager.

The Manager will automatically register the scheme for all custom resources defined under

and run all controllers under

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

mgr, err := manager.New(cfg, manager.Options{
    Namespace: namespace,

By default this will be the namespace that the operator is running in. To watch all namespaces leave the namespace option empty:

mgr, err := manager.New(cfg, manager.Options{
    Namespace: "",

Add a new Custom Resource Definition

Add a new Custom Resource Definition (CRD) API called

, with APIVersion
and Kind
$ operator-sdk add api --kind=Memcached

This will scaffold the

resource API under

Define the Memcached spec and status

Modify the spec and status of the

Custom Resource (CR) at
type MemcachedSpec struct {
    // Size is the size of the memcached deployment
    Size int32 `json:"size"`
type MemcachedStatus struct {
    // Nodes are the names of the memcached pods 
    Nodes []string `json:"nodes"`

After modifying the

file always run the following command to update the generated code for that resource type:
$ operator-sdk generate k8s

Also run the following command in order to automatically generate the CRDs:

$ operator-sdk generate crds

You can see the changes applied in


Add a new Controller

Add a new Controller to the project that will watch and reconcile the

$ operator-sdk add controller --kind=Memcached

This will scaffold a new Controller implementation under


For this example replace the generated Controller file

with the example

The example Controller executes the following reconciliation logic for each

  • Create a memcached Deployment if it doesn't exist
  • Ensure that the Deployment size is the same as specified by the
    CR spec
  • Update the
    CR status with the names of the memcached pods

The next two subsections explain how the Controller watches resources and how the reconcile loop is triggered. Skip to the Build section to see how to build and run the operator.

Resources watched by the Controller

Inspect the Controller implementation at

to see how the Controller watches resources.

The first watch is for the

type as the primary resource. For each Add/Update/Delete event the reconcile loop will be sent a reconcile
(a namespace/name key) for that
err := c.Watch(
    &source.Kind{Type: &cachev1alpha1.Memcached{}},

The next watch is for Deployments but the event handler will map each event to a reconcile

for the owner of the Deployment. Which in this case is the
object for which the Deployment was created. This allows the controller to watch Deployments as a secondary resource.
err := c.Watch(
    &source.Kind{Type: &appsv1.Deployment{}},
        IsController: true,
        OwnerType:    &cachev1alpha1.Memcached{}},

Reconcile loop

Every Controller has a Reconciler object with a

method that implements the reconcile loop. The reconcile loop is passed the
argument which is a Namespace/Name key used to lookup the primary resource object,
, from the cache:
func (r *ReconcileMemcached) Reconcile(request reconcile.Request) (reconcile.Result, error) {
    // Lookup the Memcached instance for this reconcile request
    memcached := &cachev1alpha1.Memcached{}
    err := r.client.Get(context.TODO(), request.NamespacedName, memcached)

For a guide on Reconcilers, Clients, and interacting with resource Events, see the Client API doc.

Build and run the operator

Before running the operator, the CRD must be registered with the Kubernetes apiserver:

$ kubectl create -f deploy/crds/cache.example.com_memcacheds_crd.yaml

Once this is done, there are two ways to run the operator:

  • As a Deployment inside a Kubernetes cluster
  • As Go program outside a cluster

1. Run as a Deployment inside the cluster

Build the memcached-operator image and push it to your registry. The following example uses as the registry.

$ operator-sdk build
$ sed -i 's|REPLACE_IMAGE||g' deploy/operator.yaml
$ docker push

Note If you are performing these steps on OSX, use the following

command instead:
$ sed -i "" 's|REPLACE_IMAGE||g' deploy/operator.yaml

The above command will replace the string

with the
built above. Afterwards, verify that your
file was updated successfully.
serviceAccountName: memcached-operator
- name: memcached-operator
  # Replace this with the built image name
  - memcached-operator
  imagePullPolicy: Always

IMPORTANT: Ensure that your cluster is able to pull the image pushed to your registry.

The Deployment manifest is generated at

. Be sure to update the deployment image as shown above since the default is just a placeholder.

Setup RBAC and deploy the memcached-operator:

$ kubectl create -f deploy/service_account.yaml
$ kubectl create -f deploy/role.yaml
$ kubectl create -f deploy/role_binding.yaml
$ kubectl create -f deploy/operator.yaml

NOTE: To apply the RBAC you need to be logged in

. (E.g. By using for OCP:
oc login -u system:admin.

Verify that the

Deployment is up and running:
$ kubectl get deployment
memcached-operator       1         1         1            1           1m

Verify that the

pod is up and running:
$ kubectl get pod
NAME                                  READY     STATUS    RESTARTS   AGE
memcached-operator-7d76948766-nrcp7   1/1       Running   0          44s

IMPORTANT: Ensure that you built and pushed the image, and updated the


Verify that the operator is running successfully by checking its logs.

$ kubectl logs memcached-operator-7d76948766-nrcp7
{"level":"info","ts":1580855834.104447,"logger":"cmd","msg":"Operator Version: 0.0.1"}
{"level":"info","ts":1580855834.1044931,"logger":"cmd","msg":"Go Version: go1.13.6"}
{"level":"info","ts":1580855834.104505,"logger":"cmd","msg":"Go OS/Arch: linux/amd64"}
{"level":"info","ts":1580855834.1045163,"logger":"cmd","msg":"Version of operator-sdk: v0.15.1"}
{"level":"info","ts":1580855834.1049826,"logger":"leader","msg":"Trying to become the leader."}
{"level":"info","ts":1580855834.4423697,"logger":"leader","msg":"No pre-existing lock was found."}
{"level":"info","ts":1580855834.447401,"logger":"leader","msg":"Became the leader."}
{"level":"info","ts":1580855834.7494223,"logger":"controller-runtime.metrics","msg":"metrics server is starting to listen","addr":""}
{"level":"info","ts":1580855834.7497423,"logger":"cmd","msg":"Registering Components."}
{"level":"info","ts":1580855835.3955405,"logger":"metrics","msg":"Metrics Service object created","Service.Name":"memcached-operator-metrics","Service.Namespace":"default"}
{"level":"info","ts":1580855835.7000446,"logger":"cmd","msg":"Could not create ServiceMonitor object","error":"no ServiceMonitor registered with the API"}
{"level":"info","ts":1580855835.7005095,"logger":"cmd","msg":"Install prometheus-operator in your cluster to create ServiceMonitor objects","error":"no ServiceMonitor registered with the API"}
{"level":"info","ts":1580855835.7007008,"logger":"cmd","msg":"Starting the Cmd."}
{"level":"info","ts":1580855835.7014875,"logger":"controller-runtime.manager","msg":"starting metrics server","path":"/metrics"}
{"level":"info","ts":1580855835.702304,"logger":"controller-runtime.controller","msg":"Starting EventSource","controller":"memcached-controller","source":"kind source: /, Kind="}
{"level":"info","ts":1580855835.803201,"logger":"controller-runtime.controller","msg":"Starting EventSource","controller":"memcached-controller","source":"kind source: /, Kind="}
{"level":"info","ts":1580855835.9041016,"logger":"controller-runtime.controller","msg":"Starting Controller","controller":"memcached-controller"}
{"level":"info","ts":1580855835.9044445,"logger":"controller-runtime.controller","msg":"Starting workers","controller":"memcached-controller","worker count":1}

The following error will occur if your cluster was unable to pull the image:

$ kubectl get pod
NAME                                  READY     STATUS             RESTARTS   AGE
memcached-operator-6b5dc697fb-t62cv   0/1       ImagePullBackOff   0          2m

Following the logs in the error scenario described above.

$ kubectl logs memcached-operator-6b5dc697fb-t62cv
Error from server (BadRequest): container "memcached-operator" in pod "memcached-operator-6b5dc697fb-t62cv" is waiting to start: image can't be pulled

NOTE: Just for tests purposes make the image public and setting up the cluster to allow use insecure registry. ( E.g


2. Run locally outside the cluster

This method is preferred during development cycle to deploy and test faster.

Run the operator locally with the default kubernetes config file present at

$ operator-sdk run local
INFO[0000] Running the operator locally; watching namespace "default"
{"level":"info","ts":1593777657.892013,"logger":"cmd","msg":"Operator Version: 0.0.1"}
{"level":"info","ts":1593777657.892079,"logger":"cmd","msg":"Go Version: go1.14.4"}
{"level":"info","ts":1593777657.892084,"logger":"cmd","msg":"Go OS/Arch: darwin/amd64"}
{"level":"info","ts":1593777657.892087,"logger":"cmd","msg":"Version of operator-sdk: v0.18.2"}

You can use a specific kubeconfig via the flag


Create a Memcached CR

Create the example

CR that was generated at
$ cat deploy/crds/cache.example.com_v1alpha1_memcached_cr.yaml
apiVersion: ""
kind: "Memcached"
  name: "example-memcached"
  size: 3

$ kubectl apply -f deploy/crds/cache.example.com_v1alpha1_memcached_cr.yaml

Ensure that the

creates the deployment for the CR:
$ kubectl get deployment
memcached-operator       1         1         1            1           2m
example-memcached        3         3         3            3           1m

Check the pods and CR status to confirm the status is updated with the memcached pod names:

$ kubectl get pods
NAME                                  READY     STATUS    RESTARTS   AGE
example-memcached-6fd7c98d8-7dqdr     1/1       Running   0          1m
example-memcached-6fd7c98d8-g5k7v     1/1       Running   0          1m
example-memcached-6fd7c98d8-m7vn7     1/1       Running   0          1m
memcached-operator-7cc7cfdf86-vvjqk   1/1       Running   0          2m
$ kubectl get memcached/example-memcached -o yaml
kind: Memcached
  clusterName: ""
  creationTimestamp: 2018-03-31T22:51:08Z
  generation: 0
  name: example-memcached
  namespace: default
  resourceVersion: "245453"
  selfLink: /apis/
  uid: 0026cc97-3536-11e8-bd83-0800274106a1
  size: 3
  - example-memcached-6fd7c98d8-7dqdr
  - example-memcached-6fd7c98d8-g5k7v
  - example-memcached-6fd7c98d8-m7vn7

Update the size

Change the

field in the memcached CR from 3 to 4 and apply the change:
$ cat deploy/crds/cache.example.com_v1alpha1_memcached_cr.yaml
apiVersion: ""
kind: "Memcached"
  name: "example-memcached"
  size: 4

$ kubectl apply -f deploy/crds/cache.example.com_v1alpha1_memcached_cr.yaml

Confirm that the operator changes the deployment size:

$ kubectl get deployment
example-memcached    4         4         4            4           5m


Delete the operator and its related resources:

$ kubectl delete -f deploy/crds/cache.example.com_v1alpha1_memcached_cr.yaml
$ kubectl delete -f deploy/operator.yaml
$ kubectl delete -f deploy/role_binding.yaml
$ kubectl delete -f deploy/role.yaml
$ kubectl delete -f deploy/service_account.yaml
$ kubectl delete -f deploy/crds/cache.example.com_memcacheds_crd.yaml

Reference implementation

The above walkthrough follows a similar implementation process to the one used to produce the

in the SDK samples repo.

Manage the operator using the Operator Lifecycle Manager

NOTE: This section of the Getting Started Guide is out-of-date. We're working on some improvements to Operator SDK to streamline the experience of using OLM. For further information see, for example, this enhancement proposal. In the meantime, you might find the following documentation helpful:

The previous section has covered manually running an operator. In the next sections, we will explore using the Operator Lifecycle Manager (OLM) which is what enables a more robust deployment model for operators being run in production environments.

OLM helps you to install, update, and generally manage the lifecycle of all of the operators (and their associated services) on a Kubernetes cluster. It runs as an Kubernetes extension and lets you use

for all the lifecycle management functions without any additional tools.

NOTE: Various public, OLM-ready operator projects are available at

Generate an operator manifest

The first step to leveraging OLM is to create a Cluster Service Version (CSV) manifest. An operator manifest describes how to display, create and manage the application, in this case memcached, as a whole. It is required for OLM to function.

The Operator SDK CLI can generate CSV manifests via the following command:

$ operator-sdk generate csv --csv-version 0.0.1 --update-crds

Several fields must be updated after generating the CSV. See the CSV generation doc for a list of required fields, and the memcached-operator CSV for an example of a complete CSV.

NOTE: You are able to preview and validate your CSV manifest syntax in the CSV Preview tool.

Testing locally

The next step is to ensure your project deploys correctly with OLM and runs as expected. Follow this testing guide to deploy and test your operator.

NOTE: Also, check out some of the new OLM integrations in operator-sdk:

Promoting operator standards

We recommend running

operator-sdk scorecard
against your operator to see whether your operator's OLM integration follows best practices. For further information on running the scorecard and results, see the scorecard documentation.

NOTE: the scorecard is undergoing changes to give informative and helpful feedback. The original scorecard functionality will still be available while and after changes are made.


Hopefully, this guide was an effective demonstration of the value of the Operator Framework for building and managing operators. There is much more that we left out in the interest of brevity. The Operator Framework and its components are open source, so please feel encouraged to jump into each individually and learn what else you can do. If you want to discuss your experience, have questions, or want to get involved, join the Operator Framework mailing list.

We use cookies. If you continue to browse the site, you agree to the use of cookies. For more information on our use of cookies please see our Privacy Policy.