Note: This question was pulled from an email thread.
I’ve got question regarding failback in a scenario where I had to failover to a Hyper-V site from a VMware site. When failing over from VMware to Hyper-V the machines get converted but my question is specific to the NIC in case of a fail-back. Let’s say I have a VMXNET3 NIC in my VMware VM’s (as it’s best practice) and I have to fail-over to Hyper-V it’s get set to a Synthetic NIC, but when I fail-back will I get a VMXNET3 or E1000? When going through the Cross-Hypervisor replication data sheet I only see the E1000 when going from Hyper-V to VMware but maybe it’s different in a fail-back scenario?
On failover, ZVR translates the protected VM configuration to the recovery side hypervisor VM configuration. ZVR does not keep the original VM configuration.
The VM failback initiated on the created recovery VM (contain only the recovery hypervisor side VM configuration) Thus, all unique/unmatched VM configuration parameters from the original protected VM are lost. Current implementation for ZVR 4.0 is that all VMs created on VMWare during cross-replication FA created with the E1000 NIC.
As there could be a performance impact of failing back to an E1000 adapter you could use a script to change it back to VMXNET3 on fail-back. Utilizing the Set-NetworkAdapter VMware PowerCLI cmdlet you can set the NIC Type. https://www.vmware.com/support/developer/PowerCLI/PowerCLI41U1/html/Set-NetworkAdapter.html