The architecture of a virtual machine is vastly different from traditional on-premises environments and requires different data backup techniques. This post will explain some of the best practices for backing up virtual machines.
1. Take Incremental Backups to Improve Backup Speed
Changed Block Tracking (CBT) can improve backup speeds drastically. CBT keeps track of any storage blocks that have had their data modified since the last backup. The backup application that you deploy can query the VMkernel to find the block information that has changed and back up just the changed blocks, allowing for much quicker incremental backups.
2. Snapshots Are Not Backups
Snapshots do not copy your entire VM data. The hypervisor creates a differencing disk—a special type of virtual hard disk that has a parent/child relationship with the primary virtual hard disk. Once the differencing disk is created, all write operations are directed to the differencing disk. The primary virtual hard disk of the virtual machine remains unaltered, which makes it possible to roll the virtual machine back to that earlier point in time.
Snapshots take up additional disk space on your datastores—each snapshot can grow up to the size of the original disk. The more snapshots you have running, the greater it can impact the performance of all the VMs running on the host.
Virtual machine snapshots should never be used as a primary backup means, although they are an agreeable method of backup from a short-term perspective. To learn more about why snapshots should not be used as your disaster recovery plan, click .
3. Back Up Virtual Machines at the Virtualization Layer
When backing up traditional physical servers, it is customary to install a backup agent on the guest operating system. The backup server then contacts the agent when it is about to initiate a backup operation. This method isn’t efficient in a virtual environment as it unnecessarily consumes resources on the VM and impacts the performance of the VM and all other VMs in the host.
You should instead start backing up your virtual machines at the virtualization layer. This means using a backup application that performs image-level backups of the large .vmdk file without involving the guest OS. This will ensure your VMs get all the resources they can for their workloads.
4. Copy Your Backups to a Secondary Location
Enterprise IT environments simply can’t afford to have all their backup data in a single location. Having at least one other copy of your backup data off-site ensures that disasters at the physical production location will not leave you empty handed when it comes to backups.
Ideally, the secondary backup repository should be in a different physical location or in the cloud. As long as you have network connectivity to your backup repository, you can access your backups from the secondary backup repository.
5. Encrypt Your Backups
Encryption keeps your backups secure. If someone were to get their hands on your unencrypted backup data, the backup can be restored and sensitive data in the backup can be exploited. Encrypting your backups eliminates this security risk.
6. Test Your Restoration Software Regularly
Regularly backing up your virtual machines is only the first step. You don’t want to be in a situation where you need to restore your virtual machines from a backup only to find out that your backup is corrupted.
Test your restoration software in a test environment periodically to make sure that backups are not corrupted.