Each backup vendor has its own way of doing things, so the technique used to fail back to a production server after an instant recovery isn't necessarily consistent from one product to the next. Generally speaking, the process of performing a fail back is relatively straightforward.
Once the instant recovery has been performed and the backup virtual machine (VM) is up and running, the next step is to fix any problems with the production environment. Once that task has been completed, the fail back can begin.
There are two main steps in the fail back process:
- The first can be thought of as a restoration. This essentially means copying the data from the backup virtual machine to the production environment. This is similar to a normal restoration, except the backup virtual machine remains accessible while it occurs.
- Determine what happens to data that has accumulated since the time the instant recovery took place. When an instant recovery is performed, the backup VM remains read-only. All write operations are typically directed to a differencing disk so that the backup copy of the VM remains in a pristine condition. When it is time to perform a fail back, the write operations that have occurred since the instant recovery can be merged into the newly restored production environment.
Two common mistakes can happen during the fail back process. You can accidentally delete the write operations that have occurred since the time of the instant recovery rather than merge them into the production VM. The other mistake is to forget to perform the fail back, and leave the virtual machine running in the backup environment.
How instant recovery works
Quick virtual machine backup methods for almost instantaneous recovery
Perform safer, more efficient VM backups