Volumes#

Volumes offer a high-performance alternative to cloud buckets for storing data. Unlike buckets, volumes are limited to a single region and cannot be accessed across regions and clouds.

Benefits of using volumes:

  • Performance: Volumes are generally faster than object stores, and SkyPilot lets you choose from different storage classes based on your performance requirements.

  • Data persistence: Volumes can persist data independently of task life cycles, making it easy to share data between different tasks (e.g., datasets, caches) or preserve results.

  • Size control: You can set volume size limits to manage costs and limit storage usage.

SkyPilot supports creating and managing volumes directly through the sky CLI and the web dashboard.

Supported volume types:

Volumes on Kubernetes#

In Kubernetes clusters, PVCs (Persistent Volume Claims) request and bind to PV (Persistent Volume) resources. These persistent volumes can be backed by various storage backends, including block storage solutions (AWS EBS, GCP Persistent Disk) and distributed file systems (JuiceFS, Nebius shared file system, AWS EFS, GCP Filestore), etc.

SkyPilot supports creating and managing PVC volumes in Kubernetes clusters through three commands:

  • sky volumes apply: Create a new volume

  • sky volumes ls: List all volumes

  • sky volumes delete: Delete a volume

Note

Volumes are shared across users on a SkyPilot API server. A user can mount volumes created by other users. This is useful for sharing caches and data across users.

Quickstart#

  1. Prepare a volume YAML file:

    # volume.yaml
name: new-pvc
type: k8s-pvc
infra: kubernetes  # or k8s or k8s/context
size: 10Gi
labels:
  key: value
config:
  namespace: default  # optional
  storage_class_name: csi-mounted-fs-path-sc  # optional
  access_mode: ReadWriteMany  # optional

  2. Create the volume with sky volumes apply volume.yaml:

    $ sky volumes apply volume.yaml
Proceed to create volume 'new-pvc'? [Y/n]: Y
Creating PVC: new-pvc-73ec42f2-5c6c4e

  3. Mount the volume in your task YAML:

    # task.yaml
volumes:
  /mnt/data: new-pvc  # The volume new-pvc will be mounted to /mnt/data

run: |
  echo "Hello, World!" > /mnt/data/hello.txt

Managing volumes#

List all volumes with sky volumes ls:

$ sky volumes ls
NAME     TYPE     INFRA                         SIZE  USER   WORKSPACE  AGE   STATUS  LAST_USE     USED_BY
new-pvc  k8s-pvc  Kubernetes/nebius-mk8s-vol    1Gi   alice  default    8m    IN_USE  <timestamp>  <cluster_name>

Tip

Use -v to view detailed information about a volume.

$ sky volumes ls -v
NAME     TYPE     INFRA                         SIZE  USER   WORKSPACE  AGE   STATUS  LAST_USE             USED_BY   NAME_ON_CLOUD              STORAGE_CLASS           ACCESS_MODE
new-pvc  k8s-pvc  Kubernetes/nebius-mk8s-vol    1Gi   alice  default    8m    IN_USE  2025-06-24 10:18:32  training  new-pvc-73ec42f2-5c6c4e    csi-mounted-fs-path-sc  ReadWriteMany

Delete a volume with sky volumes delete:

$ sky volumes delete new-pvc
Proceed to delete volume 'new-pvc'? [Y/n]: Y
Deleting PVC: new-pvc-73ec42f2-5c6c4e

If the volume is in use, it will be marked as IN_USE and cannot be deleted.

You can also check the volumes in the SkyPilot dashboard.

SkyPilot volumes

Filesystem volume examples#

This section demonstrates how to configure and use distributed filesystems as SkyPilot volumes. We’ll cover options like JuiceFS (a cloud-native distributed filesystem) and Nebius shared file system (a high-performance shared storage solution).

To use JuiceFS as a SkyPilot volume:

  1. Install the JuiceFS CSI driver on your Kubernetes cluster. Follow the official installation guide for detailed instructions.

  2. Verify the driver installation - Confirm that the juicefs-sc storage class has been created successfully:

$ kubectl get storageclass
NAME           PROVISIONER         RECLAIMPOLICY   VOLUMEBINDINGMODE   ALLOWVOLUMEEXPANSION   AGE
juicefs-sc     csi.juicefs.com     Retain          Immediate           false                  10m

Note

If the juicefs-sc storage class is not available, refer to the JuiceFS storage class creation guide to set it up.

  1. Create a SkyPilot volume for JuiceFS with a volume YAML:

# juicefs-volume.yaml
name: juicefs-pvc
type: k8s-pvc
infra: k8s
size: 100Gi
config:
  storage_class_name: juicefs-sc
  access_mode: ReadWriteMany

$ sky volumes apply juicefs-volume.yaml

  1. Mount the volume to SkyPilot task in your SkyPilot YAML:

# task.yaml
num_nodes: 2

volumes:
  # Mount the JuiceFS volume to /mnt/data across all nodes
  /mnt/data: juicefs-pvc

run: |
  # Verify the volume is mounted and accessible
  df -h /mnt/data
  ls -la /mnt/data

# Launch the cluster with the JuiceFS volume
$ sky launch -c juicefs-cluster task.yaml

To use Nebius shared file system as a SkyPilot volume:

  1. Set up the Nebius filesystem infrastructure by following the official documentation:

  2. Verify the storage class - Confirm that the csi-mounted-fs-path-sc storage class has been created:

$ kubectl get storageclass
NAME                     PROVISIONER                    RECLAIMPOLICY   VOLUMEBINDINGMODE      ALLOWVOLUMEEXPANSION   AGE
csi-mounted-fs-path-sc   mounted-fs-path.csi.nebius.ai  Delete          WaitForFirstConsumer   false                  10m

  1. Create a SkyPilot volume for Nebius file system with a volume YAML:

# nebius-volume.yaml
name: nebius-pvc
type: k8s-pvc
infra: k8s
size: 100Gi
config:
  storage_class_name: csi-mounted-fs-path-sc
  access_mode: ReadWriteMany

$ sky volumes apply nebius-volume.yaml

  1. Mount the volume to SkyPilot task in your SkyPilot YAML:

# task.yaml
num_nodes: 2

volumes:
  # Mount the Nebius shared filesystem to /mnt/data across all nodes
  /mnt/data: nebius-pvc

run: |
  # Verify the volume is mounted and accessible
  df -h /mnt/data
  ls -la /mnt/data

# Launch the cluster with the Nebius volume
$ sky launch -c nebius-cluster task.yaml

Volumes on GCP#

Note

GCP volume support is currently in development, and will be updated to use the sky volumes commands.

Volumes on GCP are specified using the file_mounts field in a SkyPilot task.

There are three ways to mount volumes:

  1. Mount an existing volume

  2. Create and mount a new network volume (reattachable)

  3. Create and mount a new instance volume (temporary)

To mount an existing volume:

  1. Ensure the volume exists

  2. Specify the volume name using name: volume-name

  3. Specify the region or zone in the resources section to match the volume’s location

file_mounts:
  /mnt/path:
    name: volume-name
    store: volume
    persistent: true

resources:
  # Must specify cloud, and region or zone.
  # These need to match the volume's location.
  cloud: gcp
  region: us-central1
  # zone: us-central1-a

To create and mount a new network volume:

  1. Specify the volume name using name: volume-name

  2. Specify the desired volume configuration (disk_size, disk_tier, etc.)

file_mounts:
  /mnt/path:
    name: new-volume
    store: volume
    persistent: true  # If false, delete the volume when cluster is downed.
    config:
      disk_size: 100  # GiB.

resources:
  # Must specify cloud, and region or zone.
  cloud: gcp
  region: us-central1
  # zone: us-central1-a

SkyPilot will automatically create and mount the volume to the specified path.

To create and mount a new instance volume (temporary disk; will be lost when the cluster is stopped or terminated):

file_mounts:
  /mnt/path:
    store: volume
    config:
      storage_type: instance

resources:
  # Must specify cloud.
  cloud: gcp

Note that the name and config.disk_size fields are unsupported, and will be ignored even if specified.

SkyPilot will automatically create and mount the volume to the specified path.

Configuration options#

Here’s a complete example showing all available configuration options for GCP volumes:

file_mounts:
  /mnt/path:
    store: volume

    # Name of the volume to mount.
    #
    # Required for network volume, ignored for instance volume.  If the volume
    # doesn't exist in the specified region, it will be created in the region.
    name: volume-name

    # Source local path.
    #
    # Do not set if no need to sync data from local to volume.  If specified,
    # the data will be synced to the /mnt/path/data directory.
    source: /local/path

    # If set to false, the volume will be deleted when the cluster is downed.
    # Default: false
    persistent: false

    config:
      # Size of the volume in GiB. Ignored for instance volumes.
      disk_size: 100

      # Type of the volume, either 'network' or 'instance'.
      # Default: 'network'
      storage_type: network

      # Tier of the volume, same as `resources.disk_tier`.
      # Default: best
      disk_tier: best

      # Attach mode, either 'read_write' or 'read_only'.
      # Default: read_write
      attach_mode: read_write

See YAML spec for volumes for more details.