Use Libvirt Storage Pool Methods to Migrate Libvirt Volumes

Currently, the libvirt driver only uses SSH (rsync or scp) to do cold migrations and resizes on non-shared storage. This requires SSH permissions between compute nodes, which is problematic for a number of reasons. Instead we can use the methods built in to libvirt’s storage pool API to do migrations.

NOTE: this proposal requires Use libvirt storage pools

Problem description

The primary issue is that, currently, the Nova libvirt driver requires SSH access between compute nodes to perform cold migrations and resizes on non-shared storage. This presents several issues:

  • From a security perspective, providing SSH access between compute nodes is sub-optimal. Giving compute nodes SSH access could allow a compromised node to compromise other nodes and potentially inflict harm on a cloud.

  • From a setup perspective, it adds several extra steps to a setup: System administrators, or their setup tools, configure libvirt to use a secure communication channel between source and target node. This could be the use of TLS for example. They must also generate a keypair for each compute node, and upload the public key to each of the other compute nodes.

Use Cases

Deployer: This allows the deployer to not have to set up SSH access between compute nodes while still supporting non-shared-storage resizes and cold migrations, as long as the deployer is using the new libvirt storage pools image backend.

Proposed change

The functionality in this blueprint would only be used when the deployer is using the new libvirt storage pool image backend, and has enabled the [libvirt]vol_upload_migration option.

At migration time, a new volume would be created in the destination node’s storage pool, and the methods virStorageVolDownload and virStorageVolUpload would be used to stream the contents of the disk between compute nodes (

In order to ensure secure migrations, libvirt should be configured to use one of the various forms of authentication and encryption that it supports, such as Kerberos (via SASL – or TLS client certificates (

To enable migration of suspended instances virDomainSave() will be used to save instance memory instead of virDomainManagedSave(). This will enable control of the file location, so it could be created in a directory based storage pool. This will mean you can use the storage pool APIs to upload and download that data during cold migration of suspended instances


  • Requiring shared storage for cold migrations and resizes: there are many OpenStack users who would like to be able to perform cold migrations and resizes without having shared storage.

  • Just setting up SSH keys between compute nodes: see the problem description

  • Temporarily provisioning SSH keys for the duration of the migration: While this somewhat mitigates the security issue and remove the extra setup steps, it still provides a window where the compute nodes are vulnerable. It would be much harder to secure, and would require having Nova be able to securely generate SSH keys.

  • Using an rsync daemon: People seemed to be averse to requiring an rsync daemon. Additionally, rsync daemon connections are not encryptable out of the box, although they can be run over stunnel.

Data model impact


REST API impact


Security impact

While this change does require two compute nodes’ libvirt daemons to connect to each other, such a process is already done by live migration. While the disks would no longer be encrypted by SSH while transfering, system administrators could simply secure the libvirt connections instead ( Libvirt supports TLS for encryption with x509 client certificates for authentication, as well as SASL for GSSAPI/Kerberos encryption and authentication.

Notifications impact


Other end user impact


Performance Impact

There are a couple potential performance issues:

  • Rsync can compress the contents to be transfered, although nova does not use this option. However, it does do efficient sparse handling by default.

  • Libvirt does not do compression or handle sparse files efficiently. A libvirt bug has been filed to request better sparse handling

  • In either case the use of compression is not recommended due to the impact on source node utilization.

  • The current rsync implementation will remain the default until the libvirt performance issues are addressed.

Other deployer impact

In order for the new method to work, deployers would have to enable the libvirt daemon on each compute node to listen for remote libvirt connections (if live migrations are enabled, this has already been done). Such connections must be secured as noted in Security Impact.

Additionally the configuration option :code: [libvirt]vol_upload_migration would need to be enabled. This is especially important until we can guarantee the performance is on par with the current SSH/rsync functionality (when disabled, we would fall back to the current SSH/rsync-based functionality).

Developer impact




Primary assignee:


Other contributors:


Work Items

  1. Implement the virStorageVolUpload/virStorageVolDownload code in the migrate_disk_and_power_off method, as an alternative to the existing calls to libvirt_utils.copy_image.

  2. Follow Up: remove the instances of SSH that are used to set up and tear down the migration (e.g. for shared storage detection). These could easily be done in a manner similar to how live migration works (having pre_migrate_host and pre_migrate_dest methods, instead of SSHing).


Use libvirt storage pools


Since this only changes how migration works under the hood, existing migration tests could simply be run again with the [libvirt]vol_upload_migration configuration option enabled on a setup where the libvirt storage pool image backend is also in use.

Enable these code paths and work with the “bleeding edge” libvirt tests which are being created to test.

Documentation Impact

For the OpenStack Security Guide, we should note that the new functionality can be used as an alternative to deploying SSH access between compute nodes, instead of having to provision SSH keys for the compute nodes, as well as provide instructions for securing remote libvirtd connections.

In the Compute Admin Guide, we should provide instructions for how to enable remote libvirtd connections (as required for libvirt live migration), as well as noting that these connections need to be secured, as per the Security Guide.

Since much of this documentation also applies to libvirt live migrations, it may be beneficial to place the instructions in a “general” section and link to it from both the libvirt cold migrations and libvirt live migrations documentation.