Robert L. Harris wrote: > We're working on a network-hardened solution and I've been picked to > figure out the solution to a problem. In a week I'll have hardware to > play with but I'm trying to figure out the answer or a plan of attack > early. Here's the setup: > > Server had 2 interfaces with IP's: > eth0: 192.168.1.1 > eth1: 192.168.2.1 > lo: 192.168.0.1 I can't imagine what would break if lo isn't using 127.* loopback address but I imagine it would be quite a bit. Do you have a purpose in not using 127.0.0.1 for the loopback device? I am guessing this is a mistake here. > We want to have the machine listening on eth0 and eth1 when both > networks are up and functional to the OSPF broadcasts. Oubound traffic > needs to come from the 192.168.0.1 (lo) address though so that return > traffic goes to 192.168.0.1, not the other subnets and the server accept > the packets for lo. > > Has anyone set up something like this which is relatively easy to > duplicate with a stock Linux box and Zebra for the OSPF? Perhaps I am very clueless but could you expand on just what you are trying to do here? Of course I am not familiar with Zebra. So if other people jump in with real answers please ignore me. Bob
Attachment:
pgpzkVEXQEvaL.pgp
Description: PGP signature