I have a "home" LAN in the 192.168.1.x subnet, where my internet facing router which is also the gateway has the IP of 192.168.1.1.
On top of that, I have a "lab" environment running on ESXi where I am using a 10.x.x.x subnet.
I am using RRAS to act as my router with a static route to the 192.168.1.0 subnet and another static route on my internet facing router to access the 10.x.x.x subnet.
This all works well when I'm on my laptop, but on my desktop I have a VPN client installed which gives me an IP address in the subnet 10.x.x.x, so any attempts to contact my own internal 10.x.x.x subnet get redirected through the VPN instead.
I know I could change it so that Windows prefers my non-VPN network adaptor, but that would mean the external traffic is also not routed through the VPN either, rendering the VPN useless.
So how can I gain access to my 10.x.x.x subnet whilst connected to the VPN and ensure that at the same time any external traffic is still going through the VPN?
Diagram of the network:
PC1 is running an Viscosity, an OpenVPN client to connect to a VPN service. The IP address from the VPN service varies but is always within the 10.x.x.x subnet and the subnet mask is always 255.255.255.252.
When the OpenVPN client is disconnected I can connect to all of the VMs on the 10.x.x.x subnet from PC1, but not when the VPN is connected, as the traffic gets routed through the VPN instead of outside the tunnel.
From PC2 (which does not have a VPN client) I can always connect to all the VMs on the 10.x.x.x subnet.