Hi All,
I haven’t managed to re-create this in the lab yet but before I do I’m wondering if anyone else has noticed this before…
I have fairly standard OVPN clients on RouterOS that connect back to a RouterOS OVPN Server. These work great.
During a network interruption, the OVPN client times out and tries to reconnect every few seconds. This works great and when the network resumes the OVPN tunnel comes up again as expected without any user intervention.
However, if there is ever severe packet loss on a connection which is causing the OVPN tunnel to go down/up/down/up/down/up lots and lots the client sometimes gets stuck - it shows as online (eg “R” flag) when in fact it is down. There does not seem to be any traffic between client and server at this time, and the server end simply thinks that the tunnel is down. Disabling and re-enabling the connection fixes it.
My interim fix is to use a script that disables/enables the tunnel, however this is fairly messy.
Has anyone noticed this before or labbed it up?
Cheers,
Jono.
(The config for the client is very simple, nothing fancy here)
/interface ovpn-client
add cipher=aes256 connect-to=some.address.com name=ovpn-to-some-place password=really-secret-password profile=default-encryption user=some-secret-user