Get Started With KubeRay

Install the KubeRay operator and create your first RayCluster.

Prerequisites

• A Kubernetes cluster (kind, minikube, EKS, GKE, AKS — all work)
• kubectl configured to talk to it
• helm 3.x

Install the operator

helm repo add kuberay https://ray-project.github.io/kuberay-helm/
helm repo update
helm install kuberay-operator kuberay/kuberay-operator \
  -n kuberay-system --create-namespace --version 1.4.0

Verify:

kubectl get pods -n kuberay-system

Create a RayCluster

Use the example chart for a quick start:

helm install raycluster kuberay/ray-cluster --version 1.4.0
kubectl get rayclusters
kubectl get pods -l ray.io/cluster=raycluster-kuberay

Connect

Forward the dashboard:

kubectl port-forward service/raycluster-kuberay-head-svc 8265:8265
open http://localhost:8265

Submit a job:

ray job submit --address http://localhost:8265 -- python -c "import ray; ray.init(); print(ray.cluster_resources())"

Tear down

helm uninstall raycluster
helm uninstall kuberay-operator -n kuberay-system

Next steps

RayCluster

Long-running clusters.

RayJob

Ephemeral batch jobs.

RayService

Production model serving.