A. Many virtual environment backup solutions focus heavily on the backup side of the process. In fact, the Administrator's Guide for one popular solution is over 2,000 pages long, but only dedicates a very small percentage of those pages to the actual restore process.
This lack of detailed guidance can cause extra pain when critical VMs are down. Two types of VMs in particular have special needs that must be handled prior to booting them after a restore—Exchange servers and ADDCs.
Microsoft provides very specific guidance that must be followed in the case of Exchange servers backed up using an image-level solution. That guidance is:
- Boot the Exchange VM with its mailbox stores dismounted.
- Instruct the Exchange VSS Writer to perform a restore from an available VSS snapshot
- Mount the mailbox stores
ADDCs also have an important pre-boot requirement. After restoring the VM from an image-level backup, the ADDC VM must be first booted into Directory Services Restore Mode. Not performing this step can cause the ADDC to become isolated from the rest of the domain. While unlikely (due to recent changes in Active Directory replication logic), the potential exists that it can also create a situation known as update sequence number (USN) rollback, which can create significant problems with data in the AD database.
While these two steps seem trivial, you can miss them during the flurry of activity that occurs after a server failure. That's why some image-level backup vendors now install a temporary agent into the VM as it's being backed up. This agent ensures that the VM boots up with the proper settings (or into the correct mode) after restoration, to prevent the restore from creating further problems.