A VDI outage can have a major effect on the users' ability to do their jobs. With that in mind, you need to approach disaster recovery planning from two angles in a virtual desktop environment.
When it comes to virtual desktop infrastructure (VDI), disaster recovery (DR) planning alone is not enough. An organization that depends on VDI typically can't tolerate the infrastructure going offline. As such, its first priority must be disaster prevention rather than disaster recovery. There's always the chance, however, that a completely unexpected catastrophic event could wipe out the VDI environment, so it's still important to have disaster recovery techniques at your fingertips.
Here's how to take a two-pronged approach to DR through prevention and recovery.
DR preparation starts with prevention
Disaster prevention involves using redundancy to prevent an outage. It's important to use redundant servers, but you should keep an eye on other aspects of your infrastructure that could result in a single point of failure.
More on VDI disaster recovery planning
Using VHD backups for VDI disaster recovery
Creating a disaster recovery plan for View desktops
Backup strategies for virtual desktop infrastructure
For example, virtual desktops typically reside on shared storage. The storage array should use a form of redundancy such as RAID 10 to protect against a disk failure or a disk controller failure. It is also a good idea to use redundant connectivity to the shared storage. You wouldn't want a bad host bus adapter to cause your VDI environment to fail.
Backing up VDI servers
Disaster recovery planning for VDI is relatively easy. What helps to simplify VDI disaster recovery is the fact that the underlying servers tend to run a pretty static configuration.
Front-end servers, connection brokers and virtualization host servers are rarely reconfigured aside from routine patch management. So, a good strategy for protecting those servers is to create and test a backup copy of each server. Normally, you don't need to back up the VDI servers on a regular basis unless you're making modifications to them. If you ever have to restore a VDI server, then your patch management server should be able to bring it up to date.
Recovering virtual desktops
Not only do you have to back up servers, but you must also take into account virtual desktop backups. Your VDI project's unique configuration will determine which virtual desktop disaster recovery techniques you'll use.
Most organizations generate virtual desktops from a golden image. Usually, there's no need to back up individual virtual desktops as long as you have a backup copy of the latest golden image.
Keep in mind, however, that with Windows virtual desktops, you'll need to protect user profile data. In most cases, the user profile data is not stored directly within the virtual desktop; it's usually on a file server. Regardless of the location, it's important to back up the user profiles because they contain application-specific settings, Web browser favorites, desktop customizations and in some cases even documents -- all things the user will want back.
Other components you need to protect
As you develop a VDI DR strategy, remember that creating and testing your server and virtual desktop backups is only one aspect of good disaster recovery planning. You must also consider other factors that might stand in the way of recovery. For instance, if a disaster is severe enough to significantly affect your VDI environment, then you'll probably lose some server hardware, too. Make sure you have either spare hardware on hand or a plan for shifting workloads to alternate hardware.
Finally, be sure to adequately back up servers that impact the VDI environment but that may not be technically classified as VDI servers, including infrastructure servers such as domain controllers, DNS servers, Dynamic Host Configuration Protocol servers and enterprise certificate authorities (if you use them). VDI can't function without the services these types of servers provide.
The only way to adequately protect your VDI environment is to use a two-pronged approach. First, use redundancy to prevent a disaster from occurring in the first place. Second, make sure you have reliable backups and a solid virtual desktop disaster recovery plan you can put into place if something does happen to your environment.