VCDA Failover Operations

The Migrate Operation

The Migrate operation is utilized to migrate protected virtual machines from the source (protected) site to the destination (recovery) site as a planned migration.  When performing a planned migration, it is assumed that both the source and target sites are healthy, and that you planned to relocate the virtual machines in an orderly fashion.  Using this option will shut down the VMs at the source site as the first step in the process.  When using the Migrate operation you will still have the option to configure reverse protection once the Migrate operation completes.

The Failover Operation

The Failover operation is utilized following a disaster to recover protected virtual machines to the recovery site.  A Failover assumes that connectivity between the protected and recovery sites may be down, and thus the protected virtual machines and disks are not removed.  Unlike with the Migrate operation, the Failover operation does not attempt to power down the source site because it assumes that the source site is down.  If you attempt to use the Failover operation between two EEC sites while both sites are accessible, the system will detect that VMs with identical IPs are still online at the source site and the VMs at the recovery site will not be powered on.  Manual intervention would be needed to power off the source site VMs before powering on the recovery site VMs.  When using the Failover operation you will still have the option to configure reverse protection once the Failover operation completes.

The Test Operation

The Test operation is used to test / validate the recovery of virtual machines at the recovery site.  When using the Test operation you will want to make sure that you configure bubble networks (vCD calls these "isolated" networks) on the recovery site and use the option to override the protection settings and have the test VMs connect to your bubble networks.  When used in this fashion, the Test operation keeps the source VMs online and brings the recovery site VMs online in an isolated network so that you can validate the integrity of the VMs data on the recovery site.  While you are performing a Test, replication from the source site to the recovery site continues.  

A few important notes about the Test operation:

• It is important to make sure that your Test operations are using networks that are isolated from your production networks to avoid interference
• You will need to pre-build your bubble networks on the recovery site.  Configure networks on the recovery site using the "isolated" network type and configure the same IP settings on these networks as what is configured on the source site.