Mr.PlanB Logo
    Back to Kubernetes
    Kubernetes Disaster Recovery & Migration

    Velero Backup

    Velero is the industry standard open-source tool for backing up and restoring Kubernetes cluster resources and persistent volumes, for disaster recovery, and for migrating them.

    Core capabilities

    Disaster Recovery: Reduce time to recovery in case of infrastructure loss, data corruption, or service outage.
    Data Migration: Move cluster resources to other clusters or cloud providers.
    Cluster Replication: Replicate your production cluster to development and testing environments.

    Architecture

    You need to know how Velero components interact to debug and configure it. Click a component below to see what it does in the backup process.

    Interactive System Map

    Kubernetes Cluster

    Select a Component

    Click the diagram on the left to see what each Velero component is responsible for. The same architecture handles disaster recovery and the migration of stateful applications.

    Backup workflow simulation

    A Velero backup is a coordinated sequence of API calls, snapshot triggers, and data uploads. Use the controls to step through the lifecycle of a Backup Custom Resource Definition (CRD).

    1
    API Validation
    User submits `velero backup create`. The Kubernetes API validates the CRD.
    2
    Controller Detection
    Velero Controller notices the new Backup resource and validates configuration.
    3
    Pre-Backup Hooks
    Container hooks (e.g., `fsfreeze`) run to keep application data consistent.
    4
    Resource Collection
    Velero queries the K8s API for all resources matching the label selectors.
    5
    Snapshot & Upload
    Cloud provider plugins trigger volume snapshots. Metadata is tarballed and uploaded to Object Storage.
    STATUS:IDLE

    > Waiting for backup request...

    Performance and strategy analysis

    Your choice between Native Cloud Snapshots and File System Backup (FSB, formerly Restic) affects RTO (Recovery Time Objective) and cost.

    Recovery Time (RTO) Comparison

    Lower is better

    Native Snapshots (100GB)~2 min
    FSB/Kopia (100GB)~20 min

    Insight: Native snapshots (EBS, PD) are much faster for large volumes because the cloud provider performs them at block level, bypassing the K8s network layer.

    Storage Efficiency (Deduplication)

    Kopia/Restic vs Snapshot

    FSB with Dedup (30 days)~120GB total
    Full Snapshots (30 days)~400GB total

    Insight: FSB (File System Backup) uses deduplication. For data with small daily change rates, FSB becomes cheaper than full snapshots over time, though it is slower to restore.

    Velero Operator Toolkit

    velero backup create daily-backup-01 --include-namespaces production --ttl 24h0m0s --storage-location aws-s3-primary --snapshot-volumes=true

    Related resources

    Velero Official DocumentationVelero GitHub RepositoryKubernetes Persistent VolumesLonghorn Storage for Kubernetes