Pular para o conteúdo

Apache CloudStack

Este conteúdo não está disponível em sua língua ainda.

Apache CloudStack VM disaster recovery and configuration backup via the native CloudStack API, with block-level reads over the storage network using native Ceph RBD — no SSH access to the hypervisor hosts.

  • Native CloudStack API (control) + native Ceph RBD (data)
  • Changed-block incremental backups via native RBD snapshot diff
  • VM disaster recovery
  • Configuration backup

For RBD-backed (Ceph) storage pools the plugin connects directly to the Ceph cluster with a cephx keyring and uses librbd to snapshot and read the volume over the storage network. It does not SSH into KVM hosts or shell out to the rbd CLI. Incremental backups use the native RBD snapshot diff as their changed-block map. If the native RBD path is selected but Ceph access is not configured, the job fails with an actionable error — it never silently produces an empty backup.

  • A cephx client with read + snapshot caps on the pool(s) holding the VM disks.
  • The Storage/File daemon host must reach the Ceph monitors over the storage network and have librbd installed (librbd1 / librados2).
  • Provide the cephx credentials to the plugin via ceph_conf, ceph_keyring, and ceph_client_id (and ceph_mon_host if not discoverable from ceph_conf).

Add the plugin to a Content on the client you want to protect:

Content {
Name = "cloudstack"
Include {
Options { Plugin = "cloudstack: api_url=https://cs/api api_key=KEY secret_key=SECRET vm=* ceph_conf=/etc/ceph/ceph.conf ceph_keyring=/etc/ceph/ceph.client.backup.keyring ceph_client_id=backup" }
}
}

All parameters are space-separated key=value pairs. See the Content reference for Content options.

Restore individual files, boot the VM instantly (Instant Recovery), or restore it into another hypervisor (V2V).

See also: Restore, Instant Recovery & V2V.