config DNS failed

[apstav]

Hello,
I'm new here, i have installed forticlient 7.4 in ubuntu 24.04.
I was connected to my VPN server. after a week when i tried to reconect i had this error "config DNS failed".

these are the logs from the forticlient:
20241212 13:01:50.744 TZ=+0200 [sslvpn:INFO] main:2168 Init
20241212 13:01:50.744 TZ=+0200 [sslvpn:INFO] main:2186 VPN is running in restore DNS mode
20241212 13:01:50.750 TZ=+0200 [sslvpn:INFO] nettools:1854 More than one device with index 2 can be found
20241212 13:01:50.751 TZ=+0200 [sslvpn:INFO] nettools:1854 Device address details: local_address xx.xx.xx.xx, device_index 2, device name eth0 (read from netlink)
20241212 13:01:50.751 TZ=+0200 [sslvpn:INFO] nettools:1854 Device address details: local_address xxxx::xxx:xxxx:xxxx:xxxx, device_index 2, device name eth0 (read from netlink)
20241212 13:01:50.781 TZ=+0200 [sslvpn:DEBG] vpn_util:299 Get connection name: netplan-eth0
20241212 13:01:50.791 TZ=+0200 [sslvpn:INFO] nmtools:865 Network Manager settings backup file doesn't exist
20241212 13:01:50.791 TZ=+0200 [sslvpn:DEBG] nmtools:1200 No connections to restore
20241212 13:01:50.791 TZ=+0200 [sslvpn:DEBG] dns:302 File /etc/nm_resolv.forticlient.backup doesn't exist
20241212 13:01:50.826 TZ=+0200 [sslvpn:DEBG] vpn_util:299 List fctvpn connection: netplan-eth0
lo
br-0174ab01e3e8
docker0
br-0174ab01e3e8
docker0
20241212 13:01:50.826 TZ=+0200 [sslvpn:DEBG] dns:982 default interface restore: 1, vpn interface restore: 1
20241212 13:01:50.827 TZ=+0200 [sslvpn:DEBG] mtu:121 Restore MTU.
20241212 13:01:50.827 TZ=+0200 [sslvpn:DEBG] mtu:125 No MTU backup file was found. Skip.

 Any ideas how to solve this?
Thank you

[st225]

Thanks! This worked for me too. Using Ubuntu 24.04 + FortiClient 7.4.2.1698

In my case, after Virtual Machine Manager installation, I needed to  add created virbr* device as follows:

[keyfile]
unmanaged-devices=interface-name:docker0;interface-name:br-*;interface-name:virbr*

[federicoprato]

This sounds supercool, but it doesn't seem to work for me:
do you still get all the br-* and docker0 listed when you run resolvectl dns ?
I wonder should i remove those entries from my /etc/NetworkManager/system-connections/ completely?

[federicoprato]

This works for me: i run it when i want to shut down the VPN. It shuts it down and then resets the connection i am using removing the DNS and routes for it.

#!/bin/bash

# List current running connections and allow user to select one using fzf
connection_drop_name=$(nmcli c show --active | awk 'NR>1 {print $1}' | fzf --prompt="Select a connection to drop: ")

# Check if a connection was selected
if [ -z "$connection_drop_name" ]; then
  echo "No connection selected. Exiting."
  exit 1
fi

# Bring down the selected connection
nmcli c down "$connection_drop_name"

# List current running connections and allow user to select one using fzf
connection_name=$(nmcli c show --active | awk 'NR>1 {print $1}' | fzf --prompt="Select a connection to RESET: ")

# Check if a connection was selected
if [ -z "$connection_name" ]; then
  echo "No connection selected. Exiting."
  exit 1
fi

# Modify the connection to reset DNS settings
nmcli c modify "$connection_name" ipv6.dns ""
nmcli c modify "$connection_name" ipv4.dns ""
nmcli c modify "$connection_name" ipv6.dns-search ""
nmcli c modify "$connection_name" ipv4.dns-search ""
nmcli c modify "$connection_name" ipv4.ignore-auto-dns "no"
nmcli c modify "$connection_name" ipv6.ignore-auto-dns "no"
nmcli c modify "$connection_name" ipv4.ignore-auto-routes "no"
nmcli c modify "$connection_name" ipv6.ignore-auto-routes "no"
nmcli c modify "$connection_name" ipv4.routes ""
nmcli c modify "$connection_name" ipv6.routes ""



# Restart the network manager to apply changes
sudo systemctl restart NetworkManager