I'm deploying Azure Site Recovery for an on-premises environment running about 15-20 VMware servers (Windows and Linux).
I will use a Hub & Spoke architecture on Azure. One spoke will only be peered with the hub in real DR scenario because I will use a VNet CIDR equivalent as the on-premises. The second spoke will be peered to the hub and will be used for recovery drills.
There will be a S2S VPN connection between the on-premises and Azure environments.
I read through the deployment instructions for the configuration server and it states here that a second NIC is required if we want to keep the source IP addresses in Azure and if we want to failback.
The statement is "One NIC is connected to source machines, and the other NIC is used for Azure connectivity" but I'm not sure to understand this part. One NIC will be on the local (on-premises) network but what about the second NIC ?
If there is a VPN connection, on what VNet/subnet CIDR this second NIC has to be connected ?