In this article, we will see how to install kuberay operator in a kubernetes cluster using helm. If you are looking to simplify the deployment and management of ray clusters in kubernetes then kuberay operator would be the ideal choice for this purpose. It is a free and open source Kubernetes-native operator designed to manage and deploy Ray clusters on Kubernetes. It automates the lifecycle management of Ray clusters, making it easier to run distributed machine learning (ML), AI, and data processing workloads in Kubernetes. It seamlessly integrates with Kubernetes-native tools like Helm, CRDs, and K8s autoscalers.

 

What is Ray

Ray is a free and open-source distributed computing framework designed to scale AI, ML, and data workloads including python applications efficiently across multiple nodes. It can deploy ML models as scalable APIs and can also support real-time inference workloads.

 

How to Install KubeRay Operator in a Kubernetes Cluster Using Helm

Also Read: How to Install and Use Cert Manager and cmctl for certificate generation and renewal

Step 1: Prerequisites

a) You should have a running Kubernetes Cluster.

b) You should have helm utility installed in your System. To install Helm on linux, check 6 Easy Steps to Install Helm Kubernetes Package Manager on Linux

c) You should have access to install helm chart in your Kubernetes Cluster.

 

Step 2: Add KubeRay Repo

As kuberay operator chart is available only through kuberay repo, to install the helm chart you have to first add repo by using helm repo add kuberay https://ray-project.github.io/kuberay-helm/ command as shown below.

cyberithub@cluster:~$ helm repo add kuberay https://ray-project.github.io/kuberay-helm/
"kuberay" has been added to your repositories

 

Step 3: Update Repo

After adding the repo, update the list of charts from all added repositories by using helm repo update command as shown below. This will update the local list of Helm chart repositories by fetching the latest available chart versions from remote repositories.

cyberithub@cluster:~$ helm repo update
Hang tight while we grab the latest from your chart repositories...
...Successfully got an update from the "kuberay" chart repository
Update Complete. ⎈Happy Helming!⎈

 

Step 4: Verify repo existence

You can also search the existence of kuberay related helm charts in your locally added repositories including the development versions using helm search repo kuberay --devel command as shown below.

cyberithub@cluster:~$ helm search repo kuberay --devel
NAME                           CHART VERSION   APP VERSION   DESCRIPTION 
kuberay/kuberay-apiserver      1.3.0                         A Helm chart for kuberay-apiserver
kuberay/kuberay-operator       1.3.0                         A Helm chart for Kubernetes 
kuberay/ray-cluster            1.3.0                         A Helm chart for Kubernetes

 

Step 5: Install KubeRay Operator

You can visit GitHub page and get the latest version of kuberay operator chart to install in your kubernetes cluster. At the time of writing this article, the current available version is 1.1.0 so to install this version, run  helm install kuberay-operator kuberay/kuberay-operator --version 1.1.0 command as shown below. This should deploy kuberay operator pod in default namespace to manage ray workload.

cyberithub@cluster:~$ helm install kuberay-operator kuberay/kuberay-operator --version 1.1.0
NAME: kuberay-operator
LAST DEPLOYED: Sat Mar 1 16:27:15 2025
NAMESPACE: default
STATUS: deployed
REVISION: 1
TEST SUITE: None

 

Step 6: Check kuberay-operator pods

To verify the deployment in default namespace, run kubectl get pods command. If you see output like below then it means kuberay operator deployed successfully in your kubernetes cluster.

cyberithub@cluster:~$ kubectl get pods
NAME                                READY  STATUS   RESTARTS  AGE
kuberay-operator-78cb88fb88-2tqqr   1/1    Running  0         110s