Instant VM recovery is a feature that Veeam enhanced in V10 which enable the ability to run a failed workload directly from the backup file within a few minutes.
I’ve recently been using this feature for a different use case. Moving workloads from another location where we don’t have hypervisor access in order to do a replication job to our platform.
So, what I’m doing is a Veeam in-guest agent backup of the source vm to our Cloud Connect service, importing the backup and then performing an instant recovery of the workload.
Veeam mounts the backup file using the vPower NFS service, does a number of processes including a P2V and then publishes the vm for you to access.
I then attach it to the required network on our platform. While the server is running from its mounted backup file, the user/application owner conducts tests to make sure all services and interoperability with other systems is functional. To date, none of the testers have noticed that the workloads are running from a backup file while testing. In most cases they perform faster than when they were running in their production environment and this is thanks to the pure grunt that our Cloud Connect infrastructure delivers – and this is by design.
Once confirmed that all is good, I then initiate the final step which is when Veeam does a “Migrate to production”…. essentially a storage vMotion to its new datastore.
This process has worked flawlessly and is just an example of the many ways we can get servers from other environments onto our platform.