This resource is namespace scoped, refer to [Namespace](./namespace.md) for details.

Note: similar to [Network](./network.md), cluster deletion will wait for all dependents to be deleted first.

Example:

```bash
# delete cluster and all dependents
> kubectl delete cluster 123
> kubectl delete import import-123
```

Bad example:

```bash
# only delete the Cluster, the deletion will be pended
> kubectl delete cluster 123
```

You may use `OwnerReference` to instruct Infra API to delete the dependents automatically when deleting a Cluster.

Note: [Backup](./backup.md) outlives the cluster, thus they have independent lifecycle. This is a design problem since a Backup represents both a backup job and a backup asset, and backup assets must outlive the cluster.

Note: to ease usage, `Restore` will be automatically adopted by the target cluster and delete the target cluster will cascading delete the `Restore` job.