Michael C. Toren on Tue, 20 Jun 2000 02:00:28 -0400 (EDT) |
> Now, the only thing that I'm fuzzy on is whether or not our router will > be able to figure out all this. Basically, we'll be taking our > 192.168.1.0/24 network and divide into a 192.168.1.0/25 and a > 192.168.1.128/25 networks with the router in the former network, and the > rest of our network being in the latter. This question is due mainly to > my ignorance on how routers work. :) For the most part, a router's routing table works the same way as a unix box's. The only real change is how you go about configuring those routes. I'm a little confused about your network layout, though. You have: router <-> 192.168.1.0/25 <-> linux <-> 192.168.1.128/25 <-> end user Is that correct? If so, give the router an IP address in 192.168.1.0/25 for the ethernet link it's going to share with the linux box, have the linux box default route to the router's IP on that segment, and on the router, staticly route 192.168.1.128/25 to the linux box's IP in 192.168.1.0/25. That takes care of the "router being able to figure it out" bit. What I don't quite understand, though, is how you are going to translate the RFC1918 space into space allocated to you by your upstream? Is the router doing NAT for 192.168.1/24? Does the linux box have another IP address on the 192.168.1.0/25 ethernet segment which it will be masquerading the 192.168.1.128/25 addresses behind? -mct
|
|