Den 1. dec 2004, kl. 16:16, skrev Sam Dunham:
I've got OpenVPN 2 (latest beta) installed on a Fedora Core 3 box
(server) and a Windows XP box (client). Took me a while to get the SSL
stuff worked out, but I got it working. I can connect to the server and
ping the VPN address (10.8.0.1) and ftp to that address, etc... What I
can't do is ping the internal address of the server (192.168.0.2).
I can do that, I can't connect to Samba on the server. I've tried
manipulating the routes on the server and the client to no avail.
starting OpenVPN on the server, route returns no mention of the
subnet, which is correct. Upon starting OpenVPN, I get the following
relevent info returned from route:
Destination Gateway Genmask Flags Metric
Ref Use Iface
10.8.0.2 * 255.255.255.255 UH 0
0 0 tun0
10.8.0.0 pu.lic.ho.st 255.255.255.0 UG 0 0
This seems wrong. Are you using bridging?
Now, if I'm not crazy or a complete idiot, what I need is for 10.8.0.0
to be on tun0 and I don't need 10.8.0.2 to be routed at all on the
server side. At the very least, something ain't right. Assuming the
following, what route commands do I need to run on the server and
Server internal network: 192.168.0.0
Server internal address: 192.168.0.2
Server external address: pub.lic.ho.st
Server VPN network: 10.8.0.0
Server VPN address: 10.8.0.1
Client VPN address: 10.8.0.2
Running routed, not bridged, btw.
Then your route tabled is fscked.
add a route to the 10.8.0.0 network through the tun device
add a route to 192.168.0.0 through eth0
and let /proc/sys/ipv4/ip_forward contain 1
Openvpn-users mailing list