Restore to a new deployment

Current implementation status

The backend has a clone endpoint for this workflow:

POST /api/clusters/<sourceClusterName>/restores/clone?namespace=<namespace>

The clone service creates a new cluster from the source cluster spec and adds the KubeBlocks restore annotation kubeblocks.io/restore-from-backup. The current Admin UI Create new instance from backup option is not wired to this clone endpoint. It currently calls the regular restore endpoint using the new instance name as the target cluster name, which requires that target cluster to already exist. Treat the UI path for new-instance restore as incomplete until it is rewired to the clone endpoint.

Backend requirements

The backend clone service validates:

Source cluster exists.
Backup exists.
Backup method is full.
Backup is ready for restore.
Target cluster name does not already exist in the namespace.

The clone request body uses:

{
  "newClusterName": "<new-cluster-name>",
  "backupName": "<backup-name>",
  "backupNamespace": "<backup-namespace>",
  "volumeRestorePolicy": "Parallel"
}

volumeRestorePolicy defaults to Parallel when omitted. Serial is also accepted by the service type.

Validate

kubectl get clusters.apps.kubeblocks.io -n <namespace>
kubectl get clusters.apps.kubeblocks.io -n <namespace> <new-cluster-name> -o yaml
kubectl get pods -n <namespace>

Confirm the created cluster includes the restore annotation and reaches a healthy state.

⌘I

Get started

Install

Authentication

Deploy databases

Operations

Upgrade and uninstall

Troubleshooting

Current implementation status

Backend requirements

Validate

​Current implementation status

​Backend requirements

​Validate

Current implementation status

Backend requirements

Validate