Open Source oVirt-based VMs and how to backup them?

oVirt VMs backup strategies may be more or less effective depending on the customer’s IT infrastructure. The choice between various backup strategies:

  • Export Storage Domain,
  • Disk Attachment with Proxy VM
  • Disk Image Transfer API
  • SSH Transfer
  • Change Block Tracking

depends on the technical capabilities, the size of the organization, or the specific needs and requirements of a client. However, both small businesses and enterprise clients should strive for efficient use of resources with the guarantee of data recovery when necessary. Please take into account that the above list applies to the possibilities offered by Storware vProtect – agentless VM backup solution.

Introduction to oVirt – open source virtualization solution

“oVirt is an open-source distributed virtualization solution, designed to manage your entire enterprise infrastructure. oVirt uses the trusted KVM hypervisor and is built upon several other community projects, including libvirt, Gluster, PatternFly, and Ansible.” [source]

At the outset, it is worth noting that the oVirt project was initiated by Red Hat and is the foundation on which the Red Hat Enterprise Virtualization system was built. Moreover, Red Hat Virtualization (RHV) and Oracle Linux Virtualization Manager (OLVM) derive from oVirt in the straight line. Just for the clarification – oVirt can be implemented independently.

And it is worth noting that the change of server virtualization platform from commercial solutions to oVirt is nowadays so much easier. System architects took care of the convenient way of transferring existing virtual machines from VMware, Microsoft, and Xen environments.

However, virtual machines are still prone to corruption due to many reasons. That’s why it’s crucial for maintaining business continuity for the organization to use reliable data backup and recovery solutions. The second step is to choose the best backup strategy for your data center.

Backup and recovery strategies for oVirt based virtual machines VMs:

Backup strategy 1 | Export storage domain

This is an example of the snapshot-based backup strategy. However, oVirt doesn’t allow snapshots to be exported directly but cloned to created storage domain (to the vProtect Node staging for instance). Once the snapshot was created you had to clone the VM and later export it. Keep in mind that such additional cloning requires resources (storage space and time) and the whole VM is exported, without disk exclusion.

You also are limited to 1 export storage domain in the cluster and you need to specify which hypervisor is responsible for such export. These limitations led to deprecation of this method and it is subject for future removal.

Backup strategy 2 | Disk attachment: Proxy VM

The strategy that eliminates ‘Export storage domain’ pains like no disk exclusion ability or additional cloning. By installing the vProtect Node as Proxy VM on the cluster one can forward snapshots to a backup destination, exclude drives from a backup, and read data directly from the storage.

The drawback is that you need to install Proxy VMs (at least 1 per cluster) and that it doesn’t support incremental backups.

Backup strategy 3 | Disk Image Transfer API

This option appeared with the oVirt 4.2 release. This backup strategy is easy to set up, with an incremental backup option and disk exclusion available. There is no need to install multiple Proxy VMs. It’s the single, external Node which invokes API. However, it’s worth noting that data transfer passes via the manager, which may result in a bottleneck.

Backup strategy 4 | SSH Transfer

Since this strategy backup is done over SSH directly from the hypervisor backup transfer is vastly boosted. What’s more, there is no need to install a Node on the oVirt environment. Simply speaking SSH Transfer is an improvement to the disk image transfer API backup strategy. Actual transfer may also happen using netcat, which may further improve performance.



Backup strategy 5 | Change Block Tracking 

This strategy supports incremental backups. The application of this method became possible thanks to changes in oVirt 4.4. Changed Block Tracking (CBT) uses information about zeroed and changed blocks to reduce data size and speed up the backup process. Each incremental backup uses a checkpoint which is a point in time created after the previous backup. Thus, this strategy does not require taking snapshots. Currently it is a Technical Preview, and there are still some issues waiting to be addressed in oVirt, but this strategy is going to be recommended one in the future.

oVirt VM Backup: Summary

Virtualization has become easily accessible and therefore popular and profitable also for smaller businesses. Previously, usually, only enterprise-grade companies saw its use and financial benefits.

However, virtualization alone does not solve the problems faced by modern data center architects. Infrastructure, which often consists of hundreds or thousands of virtual machines, must be properly protected so that data lost as a result of an accident or error can be easily restored.

As in industry, in the IT sector, downtime generates financial losses and often has a negative impact on the company’s image. Therefore, backup solutions are becoming a standard and integral part of the IT infrastructure.

Finally, it is worth adding that Storware has created an add-on that allows the integration of our backup solution with the native oVirt UI. The project is available on GitHub. Want to know more – click the [LINK].

Marcin Kubacki

text written by:

Marcin Kubacki, CSA at Storware