Kubernetes has steadily evolved to simplify storage management, and the v1.36 release marks a major milestone: Volume Group Snapshots have graduated to General Availability (GA). This feature, which started as an Alpha in v1.27 and went through two Beta cycles, now offers a robust way to create crash-consistent snapshots across multiple volumes. Whether you’re protecting a multi-volume application or aiming for faster recovery, these new APIs change the game. Below, we break down the ten most important things you need to know about this GA release.
1. What Are Volume Group Snapshots?
Volume Group Snapshots allow you to capture a point-in-time copy of multiple Kubernetes volumes simultaneously. Unlike individual snapshots, a group snapshot ensures that all volumes in the set are captured at the exact same moment, preserving write-order consistency. This is invaluable for applications that span several PersistentVolumeClaims (PVCs), such as databases where data and logs reside on separate volumes. The snapshot can later be used to restore new volumes or revert existing ones to that consistent state. In essence, it provides a single, atomic “photo” of a group of storage resources.
2. GA Status: From Alpha to General Availability
The journey to GA spanned multiple Kubernetes releases. Volume group snapshots first appeared as an Alpha feature in v1.27, then progressed to Beta in v1.32, and received a second Beta iteration in v1.34. With v1.36, the feature is now considered stable and production-ready. This GA designation means the API is no longer behind a feature gate, and users can rely on its backward compatibility and long-term support. The transition to GA signals that the Kubernetes community has thoroughly tested and validated the implementation, making it safe for enterprise workloads.
3. Designed for Multi-Volume Workloads
Many stateful applications—like content management systems, analytics platforms, or distributed databases—use multiple PVCs to separate data, logs, configuration, and caches. Without group snapshots, taking consistent backups required complex orchestration, often involving application quiescence (pausing I/O). Volume Group Snapshots eliminate this complexity by letting you define a group of PVCs via a label selector and snapshot them all at once. Restoring from these snapshots ensures that your application returns to a write-order-consistent state, reducing the risk of data corruption or inconsistency after recovery.
4. Crash Consistency Without Application Quiescence
One of the biggest benefits of group snapshots is achieving crash consistency without needing to pause the application. Traditional approaches would require you to freeze the application, take snapshots sequentially, then resume—a process that can be error-prone and cause downtime. With Volume Group Snapshots, the storage system guarantees that all volume copies are taken at the same instant (or as close as the underlying storage allows). This means your application can continue running while the snapshot is created, minimizing disruption while still providing a reliable recovery point.
5. The Three Core API Kinds
Kubernetes introduced three new API objects to manage group snapshots:
- VolumeGroupSnapshot – Created by a user or automation to request a snapshot of multiple PVCs.
- VolumeGroupSnapshotContent – Created by the snapshot controller for a dynamically provisioned group snapshot; it represents the actual snapshot resource in the cluster and binds to the VolumeGroupSnapshot.
- VolumeGroupSnapshotClass – Defines the driver and parameters to use when creating group snapshots (similar to StorageClass).
These APIs work together seamlessly, with the snapshot controller handling the lifecycle and deletion protection. Users interact primarily with the VolumeGroupSnapshot object, which uses a label selector to automatically include all matching PVCs.
6. Label Selectors for Dynamic Volume Grouping
A key design choice is the use of label selectors to define which PVCs belong to a group. When creating a VolumeGroupSnapshot, you specify a selector field (just like a Kubernetes selector) that matches labels on PVCs. At snapshot creation time, the controller collects all PVCs that currently match the selector and snapshots them as a group. This dynamic approach means you don’t have to manually list PVCs—instead, you can use consistent labeling policies across your workloads. It also makes it easy to include or exclude volumes by changing labels, without modifying snapshot definitions.
7. CSI Driver Support Only
Volume Group Snapshots are exclusively supported for storage systems that implement the Container Storage Interface (CSI). This means that in-tree volume plugins and legacy flexvolume drivers cannot use this feature. As CSI has become the standard for Kubernetes storage, this limitation is expected. If your storage backend doesn’t yet support CSI group snapshots, you’ll need to check with your provider for a CSI driver that implements the VolumeGroupSnapshot controller capabilities. Most major cloud and on-premise storage vendors have already added support or are in the process of doing so.
8. Differences from Individual Volume Snapshots
While Kubernetes already had the VolumeSnapshot API for single PVC snapshots, group snapshots fill a critical gap. Individual snapshots are taken independently, meaning two volumes from the same application might capture data at different moments. If you restore both, the application state could be inconsistent. Group snapshots guarantee that all volumes are captured simultaneously, achieving write-order consistency. Furthermore, you can restore a group snapshot to a new set of PVCs—essentially cloning the entire application state—which is impossible with individual snapshots alone without manual coordination.
9. Restoring a Group Snapshot
Restoring from a volume group snapshot is straightforward. You can create new PVCs pre-populated with the snapshot data by specifying a VolumeGroupSnapshot as the data source in the PVC definition. The snapshot controller will then use the underlying storage system to provision new volumes from the snapshot content. This process creates a “rehydrated” set of volumes that are crash-consistent, ready for your application to consume. Alternatively, you can use the snapshot to restore existing volumes to a previous state, though the exact mechanism depends on the storage driver’s capabilities.
10. Getting Started and Best Practices
To start using volume group snapshots, ensure your cluster is running Kubernetes v1.36 or later and that you have a CSI driver with group snapshot support. Install the snapshot controller and the associated CRDs if they aren’t already present. Then, create a VolumeGroupSnapshotClass that references your CSI driver. For best results, apply consistent labels to all PVCs that should be snapshotted together. Use descriptive labels (e.g., app: myapp, group: database) to make selector queries easy. Finally, test your backup workflow in a non-production environment before deploying to production. Group snapshots are a powerful tool for data protection and disaster recovery, especially for complex multi-volume applications.
Conclusion
The GA of Volume Group Snapshots in Kubernetes v1.36 is a significant step forward for stateful workloads. By enabling crash-consistent, multi-volume snapshots without application quiescence, this feature simplifies backup and recovery for complex applications. The new APIs integrate seamlessly with the existing snapshot ecosystem, and the use of label selectors makes grouping a breeze. If you’re running CSI-backed storage and need consistent recovery points, now is the perfect time to adopt this feature. The Kubernetes community has paved the way—your applications can now be more resilient than ever.