Docs
Kubernetes
Storage
Volume Snapshots

Working with Volume Snapshots

In Thalassa Cloud, you can create snapshots of persistent volumes to back up your data or restore a previous state. Snapshots provide a point-in-time copy of a volume, allowing you to recover from accidental data loss, perform system rollbacks, or migrate data across environments. These snapshots are managed using the VolumeSnapshot resource, which interacts with a VolumeSnapshotClass to determine how and where snapshots are stored.

Available Snapshot Classes

A VolumeSnapshotClass defines the backend storage driver and configuration used for creating snapshots. Each Thalassa Cloud Kubernetes cluster comes with a default snapshot class, named block, which is linked to Thalassa Cloud’s block storage solution. The snapshot class ensures that snapshots are created in a consistent, storage-optimized manner.

To list available snapshot classes in your cluster, run the following command:

kubectl get volumesnapshotclass

If you are using a custom deployment of Thalassa Cloud or a specific region, the snapshot class name may differ. If no snapshot classes are available, ensure that the CSI snapshot feature is enabled and properly configured for your storage provider.

Creating a Volume Snapshot

A Persistent Volume Claim (PVC) represents a storage volume used by a workload. Before creating a snapshot, you must have a PVC that you want to snapshot. Below is an example of a PVC definition:

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: my-volume
  namespace: default
spec:
  accessModes:
  - ReadWriteOnce
  storageClassName: block
  resources:
    requests:
      storage: 1Gi

Once you have a PVC, you can create a VolumeSnapshot resource. This will generate a snapshot of the specified PVC, preserving its state at the moment of creation:

apiVersion: snapshot.storage.k8s.io/v1
kind: VolumeSnapshot
metadata:
  name: my-snapshot
  namespace: default
spec:
  volumeSnapshotClassName: block
  source:
    persistentVolumeClaimName: my-volume

Apply the snapshot definition using:

kubectl apply -f snapshot.yaml

Checking Snapshot Status

Once the snapshot request is submitted, you can verify its status by running:

kubectl get volumesnapshot -A

Example output:

NAMESPACE   NAME           READYTOUSE   SOURCEPVC   SNAPSHOTCLASS   CREATIONTIME   AGE
default     my-snapshot   true         my-volume   block           5m             5m

If the READYTOUSE column shows true, the snapshot is successfully created and can be used for restoring a new volume.

Restoring a Snapshot

Restoring a snapshot allows you to create a new Persistent Volume Claim (PVC) from a previously saved snapshot. This is useful for disaster recovery, cloning environments, or testing application changes without affecting the original volume.

To restore a volume from a snapshot, define a PVC that references the snapshot in its dataSource field:

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: test-snapshot-restore
  namespace: default
spec:
  storageClassName: block
  dataSource:
    name: my-snapshot
    kind: VolumeSnapshot
    apiGroup: snapshot.storage.k8s.io
  accessModes:
  - ReadWriteOnce
  resources:
    requests:
      storage: 1Gi

Apply the restore configuration:

kubectl apply -f restore.yaml

Restoration Considerations

  • Time to Restore: The process may take several minutes, depending on the snapshot size and storage backend performance.
  • Data Integrity: The restored volume retains the exact state of the original PVC at the moment the snapshot was created. This means if the filesystem on the snapshotted volume was corrupt during snapshot creation, it will also be corrupt on the restored volume.
  • Performance Impact: Some storage backends lazy-load data from the snapshot, which may cause slightly slower performance initially until all blocks are retrieved.

Additional Resources

Persistent Volumes