mirror of
https://github.com/hanzo-ml/kuberay.git
synced 2026-10-30 19:48:47 +00:00
1.6 KiB
1.6 KiB
Installation
Make sure your Kubernetes cluster and Kubectl are both at version at least 1.23.
Nightly version
export KUBERAY_VERSION=master
# Install CRDs
kubectl create -k "github.com/ray-project/kuberay/manifests/cluster-scope-resources?ref=${KUBERAY_VERSION}&timeout=90s"
# Install KubeRay operator
kubectl apply -k "github.com/ray-project/kuberay/manifests/base?ref=${KUBERAY_VERSION}&timeout=90s"
Stable version
Method 1: Install charts from Helm repository
helm repo add kuberay https://ray-project.github.io/kuberay-helm/
# Install both CRDs and KubeRay operator
helm install kuberay-operator kuberay/kuberay-operator
Method 2: Kustomize
# Install CRDs
kubectl create -k "github.com/ray-project/kuberay/manifests/cluster-scope-resources?ref=v1.0.0&timeout=90s"
# Install KubeRay operator
kubectl apply -k "github.com/ray-project/kuberay/manifests/base?ref=v1.0.0"
Observe that we must use
kubectl createto install cluster-scoped resources. The correspondingkubectl applycommand will not work. See KubeRay issue #271.
Single Namespace version
Users can use the following commands to deploy KubeRay operator in a specific namespace.
export KUBERAY_NAMESPACE=<my-awesome-namespace>
# Install CRDs (Executed by cluster admin)
kustomize build "github.com/ray-project/kuberay/manifests/overlays/single-namespace-resources?ref=v1.0.0" | envsubst | kubectl create -f -
# Install KubeRay operator (Executed by user)
kustomize build "github.com/ray-project/kuberay/manifests/overlays/single-namespace?ref=v1.0.0" | envsubst | kubectl apply -f -