Hello everyone
I try to repost my problem.
We have 13 routers with the same problem
Each router has two wan cabled and an openvpn, to connect to the openvpn , as it usually does, the connection is through a server with public ip
The openvpn VPN is normally established if I use the main wan, when it falls the failover passes on the secondary wan but if I do a traceroute to the public ip of the openvpn server I see that continue to use the main wan (which is down) to connect then the VPN does not start.
Last thing only if I put from console the main wan deactivates (and then disappears from the routing table) obviously the VPN starts on the secondary wan
Unfortunately I had already put the 4x2 options active.
The vpn log gives me these errors
Sat Sep 21 21:48:44 2024 daemon.notice openvpn(VPN_CRI)[5569]: Attempting to establish TCP connection with [AF_INET]xx.xx.xx.xx:yyy [nonblock]
Sat Sep 21 21:48:46 2024 daemon.err openvpn(VPN_CRI)[5569]: TCP: connect to [AF_INET]xx.xx.xx.xx:yyy failed: Host is unreachable
xx.xx.xx.xx:yyy is the public ip address of openvpn server
I tried to recreate your set-up, and even without configuring Flush connection, OpenVPN server is pingable after failover.
Can you try to isolate your wan sources? Disable OpenVPN service from OpenVPN client. Then, trigger failover instance, then check if successful ping is there for the secondary WAN.
Lastly, make sure that you are on the latest FW version for the device.
