Backup and Restore Workload ID
This is a Business Critical Plan feature. For more information, see our pricing plans or contact our sales team.
Prerequisites
To set up a workload-identity, you'll need:
- A self-hosted Space cluster
- Administrator access in your cloud provider
- Helm and
kubectl
About the backup and restore component
The mxp-controller component handles backup and restore workloads. It needs to
access your cloud storage to store and retrieve backups. By default, this
component runs in each control plane's host namespace.
Configuration
Verify your configuration
After you apply the configuration use kubectl to verify the service account
has the correct annotation:
kubectl get serviceaccount mxp-controller -n ${YOUR_CONTROL_PLANE_NAMESPACE} -o yaml
Verify the mxp-controller pod is running:
kubectl get pods -n ${YOUR_CONTROL_PLANE_NAMESPACE} | grep mxp-controller
Restart workload
You must manually restart a workload's pod when you add the workload identity annotations to the running pod's service account.
kubectl rollout restart deployment mxp-controller -n ${YOUR_CONTROL_PLANE_NAMESPACE}
Use cases
Configuring backup and restore with workload identity eliminates the need for static credentials in your cluster and the overhead of credential rotation. These benefits are helpful in:
- Disaster recovery scenarios
- Control plane migration
- Compliance requirements
- Rollbacks after unsuccessful upgrades
Next steps
Now that you have a workload identity configured for the backup and restore component, visit the Backup Configuration documentation.
Other workload identity guides are: