Access existing Managed Lustre instances on GKE using the Managed Lustre CSI driver

Autopilot Standard

This guide describes how you can connect to an existing Managed Lustre instance by using the Managed Lustre CSI driver. This lets you access existing Managed Lustre instances as volumes for your stateful workloads, in a controlled and predictable way.

Before you begin

Before you start, make sure that you have performed the following tasks:

Enable the Google Cloud Managed Lustre API and the Google Kubernetes Engine API.

Enable APIs

If you want to use the Google Cloud CLI for this task, install and then initialize the gcloud CLI. If you previously installed the gcloud CLI, get the latest version by running gcloud components update.
Note: For existing gcloud CLI installations, make sure to set the compute/region property. If you use primarily zonal clusters, set the compute/zone instead. By setting a default location, you can avoid errors in the gcloud CLI like the following: One of [--zone, --region] must be supplied: Please specify location. You might need to specify the location in certain commands if the location of your cluster differs from the default that you set.

For limitations and requirements, see the [CSI driver overview].
Make sure to enable the Managed Lustre CSI driver. It is disabled by default in Standard and Autopilot clusters.

Set up environment variables

Set up the following environment variables:

export CLUSTER_NAME=CLUSTER_NAME
export PROJECT_ID=PROJECT_ID
export NETWORK_NAME=LUSTRE_NETWORK
export LOCATION=ZONE

Replace the following:

CLUSTER_NAME: the name of the cluster.
PROJECT_ID: your Google Cloud project ID.
LUSTRE_NETWORK: the shared Virtual Private Cloud network where both the GKE cluster and Managed Lustre instance reside.
ZONE: the geographical zone of your GKE cluster; for example, us-central1-a.

Configure the Managed Lustre CSI driver

This section covers how you can enable and disable the Managed Lustre CSI driver, if needed.

Enable the Managed Lustre CSI driver on a new GKE cluster

To enable the Managed Lustre CSI driver when creating a new GKE cluster, follow these steps:

Autopilot

gcloud container clusters create-auto "${CLUSTER_NAME}" \
    --location=${LOCATION} \
    --network="${NETWORK_NAME}" \
    --cluster-version=1.33.2-gke.1111000 \
    --enable-lustre-csi-driver \
    --enable-legacy-lustre-port

Standard

gcloud container clusters create "${CLUSTER_NAME}" \
    --location=${LOCATION} \
    --network="${NETWORK_NAME}" \
    --cluster-version=1.33.2-gke.1111000 \
    --addons=LustreCsiDriver \
    --enable-legacy-lustre-port

Enable the Managed Lustre CSI driver on an existing GKE cluster

If you want to enable the Managed Lustre CSI driver on an existing GKE cluster, use the following command:

gcloud container clusters update ${CLUSTER_NAME} \
    --location=${LOCATION} \
    --enable-legacy-lustre-port

After the Managed Lustre CSI driver is enabled in your cluster, you might notice that your nodes were recreated and that CPU nodes appear to be using a GPU image in the Google Cloud console or CLI output. For example:

config:
  imageType: COS_CONTAINERD
  nodeImageConfig:
    image: gke-1330-gke1552000-cos-121-18867-90-4-c-nvda

This behavior is expected. The GPU image is being reused on CPU nodes to securely install the Managed Lustre kernel modules. You won't be overcharged for GPU usage.

Disable the Managed Lustre CSI driver

You can disable the Managed Lustre CSI driver on an existing GKE cluster by using the Google Cloud CLI.

gcloud container clusters update ${CLUSTER_NAME} \
    --location=${LOCATION} \
    --update-addons=LustreCsiDriver=DISABLED

Once the CSI driver is disabled, your nodes will be automatically recreated, and the Managed Lustre kernel modules will be uninstalled from your GKE nodes.

Access an existing Managed Lustre instance using the Managed Lustre CSI driver

If you already provisioned a Managed Lustre instance within the same network as your GKE cluster, you can follow these instructions to statically provision a PersistentVolume that refers to your instance.

The following sections describe the typical process for accessing an existing Managed Lustre instance by using the Managed Lustre CSI driver:

Create a PersistentVolume that refers to the Managed Lustre instance.
Use a PersistentVolumeClaim to access the volume.
Create a workload that consumes the volume.

Create a PersistentVolume

To locate your Managed Lustre instance, run the following command.

gcloud lustre instances list \
    --project=${PROJECT_ID} \
    --location=${LOCATION}

The output should look similar to the following. Before you proceed to the next step, make sure to note down the Managed Lustre instance name, filesystem, and the mountPoint fields.

capacityGib: '18000'
createTime: '2025-04-28T22:42:11.140825450Z'
filesystem: testlfs
gkeSupportEnabled: true
mountPoint: 10.90.1.4@tcp:/testlfs
name: projects/my-project/locations/us-central1-a/instances/my-lustre
network: projects/my-project/global/networks/default
perUnitStorageThroughput: '1000'
state: ACTIVE
updateTime: '2025-04-28T22:51:41.559098631Z'

Save the following manifest in a file named lustre-pv.yaml:
```
apiVersion: v1
kind: PersistentVolume
metadata:
  name: lustre-pv
spec:
  storageClassName: "STORAGE_CLASS_NAME"
  capacity:
    storage: 18000Gi
  accessModes:
    - ReadWriteMany
  persistentVolumeReclaimPolicy: Retain
  volumeMode: Filesystem
  claimRef:
    namespace: default
    name: lustre-pvc
  csi:
    driver: lustre.csi.storage.gke.io
    volumeHandle: "PROJECT_ID/LOCATION/INSTANCE_NAME"
  volumeAttributes:
    ip: IP_ADDRESS
    filesystem: FILESYSTEM
```
Replace the following:
- storageClassName: the name of your StorageClass. The value can be an empty string, but it must meet the specification of your PersistentVolumeClaim.
- volumeHandle: the identifier for this volume.
  - PROJECT_ID: the Google Cloud project ID.
  - LOCATION: the zonal location of your Lustre instance. You must specify a supported zone for the Managed Lustre CSI driver.
  - INSTANCE_NAME: the name of your Lustre instance.
- ip: the IP address of your Lustre instance. You obtain this from the mountPoint field in the output of the previous command.
- filesystem: the file system name of your Managed Lustre instance.
For the full list of fields that are supported in the PersistentVolume object, see the Managed Lustre CSI driver reference documentation.
Create the PersistentVolume by running this command:
```
kubectl apply -f lustre-pv.yaml
```

Use the PersistentVolumeClaim to access the volume

You can create a PersistentVolumeClaim resource that references the Managed Lustre CSI driver's StorageClass.

The following manifest file shows an example of how to create a PersistentVolumeClaim in ReadWriteMany access mode , which references the StorageClass you created earlier.

Save the following manifest in a file named lustre-pvc.yaml:

  kind: PersistentVolumeClaim
  apiVersion: v1
  metadata:
    name: lustre-pvc
  spec:
    accessModes:
      - ReadWriteMany
    storageClassName: "STORAGE_CLASS_NAME"
    volumeName: lustre-pv
    resources:
      requests:
        storage: STORAGE_SIZE

Replace STORAGE_SIZE with the storage size; for example, 18000Gi. It must match the specification in your PersistentVolume.

Create the PersistentVolumeClaim by running this command:
```
  kubectl create -f lustre-pvc.yaml
```

Create a workload that consumes the volume

This section shows how to create a Pod that consumes the PersistentVolumeClaim resource you created earlier.

Multiple Pods can share the same PersistentVolumeClaim resource.

Save the following manifest in a file named my-pod.yaml.

  apiVersion: v1
  kind: Pod
  metadata:
    name: my-pod
  spec:
    containers:
    - name: nginx
      image: nginx
      volumeMounts:
        - name: lustre-volume
          mountPath: /data
    volumes:
    - name: lustre-volume
      persistentVolumeClaim:
        claimName: lustre-pvc

Run the following command to apply the manifest to the cluster:
```
  kubectl apply -f my-pod.yaml
```
The Pod waits until GKE provisions the PersistentVolumeClaim before it starts running. This operation might take several minutes to complete.
Verify that the Pod is running:
```
  kubectl get pods
```
It might take a few minutes for the Pod to reach the Running state.

The output is similar to the following:
```
  NAME           READY   STATUS    RESTARTS   AGE
  my-pod         1/1     Running   0          11s
```

Use fsGroup with Managed Lustre volumes

You can change the group ownership of the root level directory of the mounted file system to match a user-requested fsGroup specified in the Pod's SecurityContext.

Troubleshooting

For troubleshooting guidance, see the Troubleshooting page in the Managed Lustre documentation.

Clean up

To avoid incurring charges to your Google Cloud account, delete the storage resources you created in this guide.

Delete the Pod and PersistentVolumeClaim.

Note: When you create the PersistentVolume with a "Retain" persistentVolumeReclaimPolicy, deleting the PersistentVolumeClaim won't remove the PersistentVolume or the underlying Managed Lustre instance.
```
kubectl delete pod my-pod
kubectl delete pvc lustre-pvc
```

Check the PersistentVolume status. After deleting the Pod and PersistentVolumeClaim, the PersistentVolume should report a "Released" state:

kubectl get pv

The output is similar to the following:

NAME        CAPACITY   ACCESS MODES   RECLAIM POLICY   STATUS     CLAIM                 STORAGECLASS   REASON   AGE
lustre-pv   18000Gi      RWX            Retain        Released   default/preprov-pvc                           2m28s

Reuse the PersistentVolume. To reuse the PersistentVolume, remove the claim reference (claimRef):

kubectl patch pv lustre-pv --type json -p '[{"op": "remove", "path": "/spec/claimRef"}]'

The PersistentVolume should now report an "Available" status, indicating its readiness to be bound to a new PersistentVolumeClaim. Check the PersistentVolume status:

kubectl get pv