Multiple Interfaces Issue

Seth Mos knuffie at xs4all.nl
Tue Dec 31 10:46:00 CST 2002


At 10:17 31-12-2002 -0500, John Reuning wrote:
>On Tue, 2002-12-31 at 09:27, bscott at ntisys.com wrote:
> > On Mon, 30 Dec 2002, at 10:44pm, john at metalab.unc.edu wrote:
> > >> 10.10.10.0    *            255.255.255.0    U      40   0        0  eth1
> > >> 10.10.10.0    *            255.255.255.0    U      40   0        0  eth1
> > >> default       10.10.10.1   0.0.0.0          UG     40   0        0  eth0
> > >
> > > Hmmm... the routing tables could be the source of the problem.  In the
> > > config [above], it looks like the machine thinks that its subnet is only
> > > on eth1, not on both interfaces.  I'm not sure which config option the
> > > redhat scripts use to set the routes.
> >
> >   Routes to directly-connected networks spring into existence automatically
> > when the interfaces are configured.  Or at least, that is the way it is
> > supposed to work, in kernel 2.4.
>
>But shouldn't there be 2 routes for 10.10.10.0, one for eth0 and one for
>eth1?  As you indicated in the earlier message, it looks like either a
>config file error or a problem with the RH network control scripts.

Have you tried configuring the interfaces with the plain ifconfig command 
on the prompt?

ifconfig eth0 10.10.10.5
ifconfig eth1 10.10.10.6
route add default gw 10.10.10.1

That should work. Another way would be to set up policy based routing, 
which is a) overkill and b) not neccesary.

Is it possible that you have the reverse path filter (used to protect 
against spoofing) switched on? This might cause "weird" behaviour on 
multihomed crossing networks.

I use the policy routing available in the linux kernel since I have a 
multihomed host that also needs the spoofing protection.

I can give you an example if you want it.

Cheers
--
Seth
It might just be your lucky day, if you only knew.




More information about the Linux-PowerEdge mailing list