OpenVPN problems on WAN failover

Hello everybody,

I’m running a PBX behind a OpnSense that is running a OpenVPN-server.
I have a Mikrotik running as a VPN-client to that server.

When the VPN is connected everything works as expected but I can see a weird issue:
The Mikrotik VPN-client is given the address 10.20.30.0/24 instead of eg. 10.20.30.2/24?
Perhaps it’s just some visual bug, everything works but it’s just annoying. I have attached an image of the address from the Mikrotik.
If I have made a mistake on the config of the VPN I’m willing to change it :slight_smile:.
Config for the VPN in the Mikrotik:

add certificate=FREEPBX_VPN cipher="replaced" connect-to=x.x.x.x mac-address=02:00:0E:69:92:82 name=OpenVPN user=anyuser

add action=accept chain=forward comment="ALLOW PHONE_VLAN ACCESS TO TELFIREWALL1" in-interface-list=PHONE out-interface=OpenVPN
add action=accept chain=forward comment="ALLOW FREEPBX ACCESS TO PHONE_VLAN" in-interface=OpenVPN out-interface-list=PHONE

However, when I do a failover to the secondary internet connection all communication on the VPN goes down. I can’t get ping or any other traffic through.
I can restart the OpenVPN-server but that doesn’t help but as soon as I disable and enable the VPN-client in the Mikrotik it get’s going again.
Then I start WAN1 again and the VPN goes down. I have waited about half an hour but it won’t come back up without me disabling and enabling the VPN-client in the Mikrotik.

I can solve this by scripting or similar but it would be great if there is simply something I have missed in the configuration? I would prefer to solve the root cause.
RHjkMJuSh5.png

I solved the first part of this issue by simply running Wireguard instead.

It’s frustrating though because one can do so much more with OpenVPN so I would really prefer that but Wireguard will do until then :stuck_out_tongue:.