Partial Failover
Last updated
Last updated
Partial failover is used to engage DRaaS for individual VMs in the event the production site and primary infrastructure is still available, but there is an issue with one or more VMs (eg. single host failure or OS corruption) and the affected VMs need to be temporarily booted in the DR environment.
Partial failover allows VMs booted in the DR environment to continue communicating with the internet and remaining devices in the production site through the use of a Veeam Network Extension Appliance (NEA) in each location. This is deployed during the setup of the Service Provider connection in Veeam B&R.
During a partial failover, the extension appliances establish a tunnel between each other and use proxy ARP to extend layer 2 connectivity between the production and DR sites.
VMs running in the DR environment continue to route to the internet and the production network via the NEA tunnels, providing a near seamless option for recovering individual machines to the DR environment.
Once the production infrastructure is ready to run the affected VMs again, they can be failed back to the production system with minimal downtime.
Partial failover requires Network Extension Appliances deployed in both the production and DR environments. Please ensure you have deployed at least 1 appliance during setup of the Service Provider or partial failover will not be successful.
See the Network Extension section of Getting Started for more information
From the on-prem Veeam B&R console, select Replicas from the navigation tree
Right click the VM you wish to failover and select Failover Now
Select a restore point and click through to Finish
Select a restore point and click through to Finish
Wait for the failover process to complete. While this process is running, the replica VM and Network Extension Appliance will be booted and Veeam will verify connectivity between the cloud and on-prem network appliances is successfully established
If the process completed successfully, your VM should now be contactable from the existing on-prem network
