On 26 May 2004 11:23:17 -0700, Dale Martenson <(E-Mail Removed)> wrote:
> I have a dial-in server (24 line off a T1) using portslave where
> clients dial-in and get access to the network just fine using a pool
> of ip addresses on the same subnet as the server.
>
> Now there is one dial-in client that is actually a router that wishes
> to use a static ip (not on the same subnet). How do I get this to
> work? I assume I need to add a static route, but I don't know which
> ppp interface the client is going to dial-in on. Will I have to add
> something to \etc\ppp\ip-up to special case this?
Assuming you are using the router's IP as remote IP from your end, it
should end up with a host route and you just need to add the network route
from ip-up (or ip-up.local) if the remote IP ($5) is 172.11.1.1:
if [ "$5" = "172.11.1.1" ] ; then
echo "Router connecting"
echo "Do /sbin/route -net whatever-net&netmask dev $1"
fi
> Example:
>
> My server 192.168.1.2 with a gateway at 192.168.1.1
>
> Router with static ip 172.11.1.1
>
> When the router connects, it can ping 192.168.1.2, but nothing else.
How would any other boxes on LAN know to route 172.11.1.1 (which is NOT a
private IP range, but currently unassigned) to 192.168.1.2 instead of
their default gateway 192.168.1.1? If the gateway has a suitable OS,
maybe you could set up a pair of alias IPs between the gateway and your
server, so the default gateway could route that network to your server
(instead of internet where it gets lost). If that 172. net wants to
access the internet through you, your gateway would also need to
masquerade it.
--
David Efflandt - All spam ignored
http://www.de-srv.com/