Skip to content

vm_file_restore.md

Latest commit

 

History

History
231 lines (184 loc) · 6.14 KB

vm_file_restore.md

File metadata and controls

231 lines (184 loc) · 6.14 KB

VM File Restore Configuration

Enable VM File Restore

The VM File Restore feature allows you to perform granular file-level restore from VM backups without restoring the entire virtual machine. This is particularly useful when you only need to recover specific files or directories from a backed-up VM.

Prerequisites

Before enabling VM File Restore, ensure the following prerequisites are met:

  1. OpenShift Virtualization: KubeVirt must be installed and configured in your OpenShift cluster
  2. Required Plugins: The following Velero plugins must be configured in your DPA:
    • kubevirt - Provides VM backup and restore capabilities
    • openshift - Required for OpenShift-specific functionality

Basic Configuration

To enable VM File Restore, add the vmFileRestore section to your DataProtectionApplication CR:

apiVersion: oadp.openshift.io/v1alpha1
kind: DataProtectionApplication
metadata:
  name: dpa-sample
  namespace: openshift-adp
spec:
  configuration:
    velero:
      defaultPlugins:
        - openshift
        - aws
        - kubevirt  # Required for VM file restore

  # Enable VM File Restore
  vmFileRestore:
    enable: true

  backupLocations:
    - name: default
      velero:
        provider: aws
        default: true
        objectStorage:
          bucket: my-backup-bucket
          prefix: velero
        config:
          region: us-east-1

Custom Resource Configuration

You can specify custom resource limits and requests for the VM File Restore controller:

apiVersion: oadp.openshift.io/v1alpha1
kind: DataProtectionApplication
metadata:
  name: dpa-sample
  namespace: openshift-adp
spec:
  configuration:
    velero:
      defaultPlugins:
        - openshift
        - aws
        - kubevirt

  vmFileRestore:
    enable: true
    resources:
      limits:
        cpu: "1"
        memory: 256Mi
      requests:
        cpu: 100m
        memory: 128Mi

  backupLocations:
    - name: default
      velero:
        provider: aws
        default: true
        objectStorage:
          bucket: my-backup-bucket
          prefix: velero
        config:
          region: us-east-1

Default Resources (if not specified):

  • CPU Limit: 500m
  • Memory Limit: 128Mi
  • CPU Request: 10m
  • Memory Request: 64Mi

Image Overrides

For disconnected environments or testing, you can override the VM File Restore controller images using unsupportedOverrides:

apiVersion: oadp.openshift.io/v1alpha1
kind: DataProtectionApplication
metadata:
  name: dpa-sample
  namespace: openshift-adp
spec:
  configuration:
    velero:
      defaultPlugins:
        - openshift
        - aws
        - kubevirt

  vmFileRestore:
    enable: true

  unsupportedOverrides:
    vmFileRestoreControllerImageFqin: "my-registry.io/oadp-vm-file-restore:v1.0.0"
    vmFileRestoreAccessImageFqin: "my-registry.io/oadp-vmfr-access:v1.0.0"
    vmFileRestoreSSHImageFqin: "my-registry.io/oadp-vmfr-access-sshd:v1.0.0"
    vmFileRestoreBrowserImageFqin: "my-registry.io/oadp-vmfr-access-filebrowser:v1.0.0"

  backupLocations:
    - name: default
      velero:
        provider: aws
        default: true
        objectStorage:
          bucket: my-backup-bucket
          prefix: velero
        config:
          region: us-east-1

Using VM File Restore

Once enabled, the VM File Restore controller will be deployed in the OADP namespace. You can then use the following Custom Resources:

1. VirtualMachineBackupsDiscovery (VMBD)

Discover available VM backups:

apiVersion: oadp.openshift.io/v1alpha1
kind: VirtualMachineBackupsDiscovery
metadata:
  name: my-vm-discovery
  namespace: openshift-adp
spec:
  backupName: my-vm-backup

2. VirtualMachineFileRestore (VMFR)

Restore specific files from a VM backup:

apiVersion: oadp.openshift.io/v1alpha1
kind: VirtualMachineFileRestore
metadata:
  name: my-vm-file-restore
  namespace: openshift-adp
spec:
  backupName: my-vm-backup
  volumeName: my-vm-disk
  fileAccessMethod: ssh  # or filebrowser

File Access Methods

The VM File Restore feature supports two methods for accessing restored files:

  1. SSH Access: Connect to restored files via SSH (requires SSH client)
  2. FileBrowser: Web-based file browser interface for easy file access

Validation Requirements

When VM File Restore is enabled, OADP validates:

  • ✓ Velero configuration is present
  • ✓ kubevirt-velero-plugin is configured in defaultPlugins
  • ✓ openshift-velero-plugin is configured in defaultPlugins
  • ✓ Only one DPA instance across the cluster has VM File Restore enabled

Disabling VM File Restore

To disable VM File Restore, set enable to false:

apiVersion: oadp.openshift.io/v1alpha1
kind: DataProtectionApplication
metadata:
  name: dpa-sample
  namespace: openshift-adp
spec:
  configuration:
    velero:
      defaultPlugins:
        - openshift
        - aws
        - kubevirt

  vmFileRestore:
    enable: false  # Disables VM file restore

The VM File Restore controller deployment will be automatically removed when disabled.

Troubleshooting

VM File Restore controller not deploying:

  • Check that both kubevirt and openshift plugins are in defaultPlugins
  • Verify that KubeVirt is installed in your cluster
  • Check OADP operator logs for validation errors

Cannot create VirtualMachineFileRestore resources:

  • Ensure VM File Restore is enabled in the DPA
  • Verify the controller deployment is running: oc get deployment -n openshift-adp oadp-vm-file-restore-controller-manager
  • Check controller logs for errors: oc logs -n openshift-adp deployment/oadp-vm-file-restore-controller-manager

Multiple DPA error:

  • Only one DPA instance across the entire cluster can have VM File Restore enabled
  • Disable VM File Restore in other DPA instances if the error occurs

Additional Resources