Skip to main content
Backup for GKE is a fully managed Google Cloud service that lets you consistently back up and restore workloads running in Google Kubernetes Engine (GKE) clusters. By creating regular snapshots of your Kubernetes resources and persistent volumes, you can:
  • Accelerate disaster recovery and minimize downtime
  • Support CI/CD pipelines and environment cloning
  • Perform cluster upgrades and migrations safely
  • Meet Recovery Point Objectives (RPOs) and compliance requirements
The image is an overview of "Backup for GKE" on Google Cloud Platform, highlighting features like disaster recovery, CI/CD pipelines, workload cloning, and performing upgrades.

​
Enabling and Using Backup for GKE

Follow these steps to get started with Backup for GKE in your cluster:
  1. Enable the add-on
    Enable the Backup for GKE add-on on your target cluster using the Cloud Console or gcloud:
  2. Configure a Backup Plan
    Define which namespaces, workloads, or volumes you want to include. You can back up:
    • All workloads in the cluster
    • Specific namespaces or labels
    • Individual PersistentVolumeClaims (PVCs)
  3. Create a Backup
    Trigger an ad-hoc backup or schedule recurring jobs:
  4. Restore a Backup
    Target any GKE cluster with the add-on enabled. You can restore into:
    • The original cluster (overwriting existing resources)
    • A different cluster for cloning or testing
Before using Backup for GKE, ensure your IAM user or service account has the roles/gkebackup.admin role.

​
Backup for GKE Architecture

Backup for GKE consists of two primary components that work together to orchestrate backups and restores:
The image is a diagram illustrating the architecture of a backup system for Google Kubernetes Engine (GKE), showing components like tenant projects, GKE clusters, and various APIs and services.

​
Supported vs. Excluded Resources

Backup for GKE automatically captures Kubernetes manifests and the data in PersistentVolumeClaims. However, some elements are not included:
The image is a diagram titled "Backup for GKE – Considerations," showing three key components: GKE Cluster Configuration, Container Images, and External Services, all connected to GKE.
If an image is removed from its registry after you’ve backed up the manifest, restores that reference will fail. Always maintain image retention policies or mirror images in a private registry.

​
Designing a Comprehensive Recovery Strategy

To ensure full resilience and meet your RPO/RTO goals, augment Backup for GKE with:
  • Cluster Configuration Management
    Use Infrastructure as Code (IaC)—Terraform or Deployment Manager—to version and restore node pools, network settings, and IAM policies.
  • Container Image Retention
    Implement image lifecycle policies in Artifact Registry or Container Registry to prevent accidental deletion.
  • External Service Backups
    Schedule snapshots for managed databases (e.g., Cloud SQL) and configurations for external load balancers or DNS records.

Watch Video