We implement Kubernetes cluster backup with Velero — complete backups of applications, cluster resources (including CRDs) and PV data, with restore in place, to a new cluster or into a cloud environment.
What it covers
- Cluster object backup: Namespaces, Deployments, Services, ConfigMaps, Secrets and CRDs
- PV data backup — CSI snapshots (VolumeSnapshots) and file-based copies (restic/Kopia)
- Scheduled backups and TTL retention policies
- S3-compatible storage: MinIO/S3 (self-hosted), Wasabi, AWS S3, Azure Blob
- Application migrations between clusters — backup in one, restore in another (dev/stage/prod)
- Recovery drills and Disaster Recovery procedures with RTO/RPO targets
- Monitoring integration — alerts on failed backups
- Extensions: cloud provider plugins, encryption, multiple BackupStorageLocations
We keep the configuration (velero install, schedules, storage location) in a GitOps repository and test restores regularly — so recovery after an outage is real, not just theoretical.