The module lifecycle stage: Preview

What is the state-snapshotter module?

The state-snapshotter module is a critical infrastructure component in Deckhouse that provides state capture and restoration capabilities for Kubernetes resources. It allows you to create point-in-time snapshots of specific Kubernetes objects and restore them later.

What does it do?

The module provides:

  1. State Capture: Captures the current state of specific Kubernetes resources (CRDs, ConfigMap, Secret, etc.) into immutable checkpoints
  2. State Storage: Stores captured state in a distributed format with chunking, optimized for large datasets
  3. State Restoration: Retrieves and restores captured state through a native Kubernetes API
  4. Automatic Lifecycle Management: Automatically manages checkpoint storage and cleanup using TTL-based retainers

Why is it needed?

The state-snapshotter module is essential for:

  • Backup and Restore Operations: Create backups of application state before important operations (upgrades, migrations, etc.)
  • Disaster Recovery: Maintain point-in-time snapshots for disaster recovery scenarios
  • State Migration: Capture state from one cluster and restore it in another
  • Audit and Compliance: Maintain historical records of resource states for audit purposes
  • Integration with Other Modules: Used by other Deckhouse modules (e.g., virtualization, backup) to capture and restore application state

Why you should NOT disable it

⚠️ Important: The state-snapshotter module should NOT be disabled because:

  1. Critical Infrastructure: It’s a foundational component used by other Deckhouse modules for state management
  2. No Performance Impact: The module only activates when explicitly requested via ManifestCaptureRequest resources - it doesn’t run background processes
  3. Resource Efficient: Uses minimal resources and only processes explicit capture requests
  4. Required for Operations: Many Deckhouse operations (backups, snapshots) depend on this module
  5. Automatic Cleanup: Automatically manages its own resources and cleans up old checkpoints based on TTL