Microsoft’s Hyper-V is quickly becoming a common platform for enterprise datacenters looking to virtualize infrastructure. With the latest version of Hyper-V technology with Windows Server 2016, the maturity of the platform is no longer questioned. With this being aid, organizations using Microsoft Hyper-V for enterprise virtualization must consider many different things when architecting a backup solution for Microsoft Hyper-V environments. When it comes to making sure data is protected, there is no substitute for taking proper backups of environments even if organizations are using replication, or other means to have copies of data in various locations. Let’s take a look at Hyper-V backup considerations and how organizations can make sure they have an effective and efficient data protection plan covering their Hyper-V infrastructure.
Hyper-V Backup Considerations
There are several things that need to be considered when thinking about backing up Microsoft Hyper-V environment. Among the considerations we need to make include the following:
- Backing up at the host or virtual machine level?
- Virtual machine virtual hardware configuration
- Resilient Change Tracking
- Application aware backups
- Saved State vs Child VM snapshots
- Hyper-V Integration services
Hyper-V Backup at Host or Guest level?
When interacting with virtual machines for backup purposes, one of the first considerations we need to make is whether we are going to backup the guest virtual machine from the host level or from within the guest operating system. There are certainly advantages in backing up production workloads at the host level. Backing up virtual machines at the host level means we are performing backups by interacting with the Hyper-V host itself. This is generally desirable from a backup management perspective since we interact with the host instead of multiple virtual machines. Backup solutions that are able to perform image-based backups from the host level also capture the virtual machine hardware and settings. This is highly desirable from a restorability standpoint since everything required to restore the virtual machine is included in the backup.
Often if using a legacy backup solution that is not able to interact with the Hyper-V host directly, organizations may only have the option of performing backups with guest agents installed. This approach can become quite inefficient from a manageability standpoint since agents have to be installed, updated, etc. Additionally, the virtual machine hardware configuration and settings are not captured when using an in guest agent.
Virtual Machine Virtual Hardware Configuration
There is certainly considerations to be made with Hyper-V backups and the type of virtual hardware a virtual machine is configured with as well as the type of virtual disk the virtual machine is running on. In Windows Server 2012 R2 host level backups were drastically improved in their ability to scale. This is due to the VSS snapshot being moved from the host side into the virtual machine itself. However, this introduced a limitation on host level backups for virtual machines that didn’t have a SCSI controller configured. The SCSI controller is required due to the hot add process of the snapshot. This can present a limitation of host level backups with certain virtual machines not configured with SCSI controllers.
Additionally, to be considered is whether or not a virtual machine is running with a VHD disk or VHDX disk configured. This comes into play with Resilient Change Tracking or RCT which we will discuss below and the ability to implement RCT. A requirement of RCT is the virtual machine files need to be VHDX files.
Resilient Change Tracking (RCT)
Resilient Change Tracking was introduced with Windows Server 2016 Hyper-V and it allows native “changed block tracking” to be implemented with virtual machine backups. Resilient Change Tracking is Microsoft’s implementation of the native ability to track which blocks have changed between backup iterations. After the first full backup, RCT is able to map the blocks that have changed since the last backup, allowing backups to be much more efficient than before. Previously, backup vendors had to use a proprietary filter driver to determine the blocks that had changed with Hyper-V. This was quite a bit of overhead on performance and introduced other potential issues. Now with RCT, this functionality is native to Hyper-V. As mentioned above, there are certain requirements for RCT to work such as the virtual machine running on VHDX disks. Additionally, virtual machine version needs to be version 7 or above. Running a lower VM version results in the “differencing disk” method being used as is the case with Windows Server 2012 R2.
Application Aware Backups
If you have virtual machines running in your Hyper-v environment hosting application services such as databases that need transactional consistency, you need to make sure your Hyper-V backups are “application aware”. Application aware backups make sure there is no transactional information living in memory or any pending I/O operations that haven’t been committed to disk before taking a backup. This ensures that database driven applications such as Microsoft Exchange Server and Microsoft SQL Server are backed up in a consistent state. It also makes sure there are no additional restore steps needed to bring the database up to a consistent state.
Saved State vs Child VM Snapshots
There are two methods used in Hyper-V for taking backups of virtual machines – saved state and child VM snapshots. Saved state is the method that Hyper-V uses if no other method is requested or if the virtual machine doesn’t qualify for the child VM snapshot method. The saved state method results in a slight disruption in the reachability of the Hyper-V virtual machine and is less desirable. It is quickly hibernated and then brought out of a hibernated state by the Hyper-V VSS writer. Saved state backups are not application aware which means we can’t have transactionally consistent backups of applications using this method.
The child VM snapshot method is the preferred method for taking backups in Hyper-V. The child VM method is known as an “online” backup since the virtual machine remains online during the VSS operation. It requires the virtual machine is powered on, has Hyper-V integration services installed, the VM is running basic disks, and the file system is one that supports snapshots such as NTFS.
Hyper-V Integration Services
Hyper-V Integration Services allow special interaction between the Hyper-V host and the running virtual machines. Especially when dealing with backup processes, the Hyper-V Integration Services allow backups of the guest operating system to be in a consistent state as it works in harmony with the VSS services. Having Hyper-V integration services installed and at the current version ensures backups can take advantage of the child VM snapshot method mentioned above for backups.
Thoughts
Microsoft Hyper-V offers a powerful enterprise virtualization platform especially with Windows Server 2016. Organizations will want to pay attention to the various Hyper-V backup considerations mentioned above and others that are specific and possibly unique to their individual Hyper-V environments. Additionally, utilizing a powerful and agile backup solution such as Vembu BDR Suite that leverages all the most recent capabilities found in Microsoft Hyper-V will allow consistent, effective, and efficient backups of Hyper-V environments.
Experience modern data protection with this latest Vembu BDR Suite v.3.8.0 FREE edition. Try the 30 days free trial here: https://www.bdrsuite.com/vembu-bdr-suite-download/
Got questions? Email us at: vembu-support@vembu.com for answers.
Follow our Twitter and Facebook feeds for new releases, updates, insightful posts and more.
Note: This is a Republished Article. Blog originally published on – November 13th, 2017
