Hello. I have two LAN VPN points. One with Win XP
[A] (simple workstation, 1 ethernet card), and another with Linux
[B](also with 1 ethernet card bridged with the tap one). Well. The issue is: I
want to prepare both stations to be able to connect themself
with OpenVPN and being able to connect also to other stations
without VPN, standard communication. Even between them 2 also, if
possible.
Then, in the first station (A) i did nothing
special, ip 192.168.130.2, and the second one (B): 192.168.130.1
With tap issues, A: 10.30.20.2 , and B has
no IP in its tap interface (Because of being bridged) to receive tunneled
data.
So, the problem is that.... theoretically, B will
need another virtual interface, so i added it with the simple ifconfig eth0
add 10.30.20.1. Well. Then... The VPN pings do not arrive between stations; i
think because A tries to send the ping through the tunnel and B answers
without tunneling (because of having an added interface). Could the tap0
interface have an own IP and be practicable?
So, the problem is, i supose, at the routing
table level. What could i do? Maybe establishing lower metrics to the tap
interface in (B) for the tap interface? I tried doing that, adding the
remote station (A) as a unique route with the interface tap0 as
'gateway' (route add 10.30.20.2 dev tap0). It didnt work. Should i create
another tap interface instead of 'ifconfig eth0 add xxx' ? Which is the
possible 'escape' with a bridged host/server if you want also it to receive
connections from both, tunnel and simple ipv4?
Thanks.