<< | Thread Index | >> ]    [ << | Date Index | >> ]

Subject: Re: A minor problem...
From: ewheeler,AT,kaico,DOT,com
Date: Fri, 17 Aug 2001 23:36:20 +0200
In-reply-to: <3B7D5347.4A3888@asylum.xs4all.nl>

Peter --
  I understand your concern.  You might look into unnumbered PPP
links.  Basically, the PPP link takes the address of an interface so you
don't have an additional network to route over.  Since cipcbX is a virtual
network device, it will pass frames back and fourth between it and it's
bound adapter (as I understand it).  It makes things much cleaner in the
long run.  Here is how I have our network setup and it is very fast and
stable:

Center Hub:
  eth1: internet ip: 1.2.3.4
  eth0: 192.168.1.1
  cipcb0: ip: 192.168.1.1 ptp: 192.168.10.1
  cipcb1: ip: 192.168.1.1 ptp: 192.168.11.1
  cipcb2: ip: 192.168.1.1 ptp: 192.168.12.1
  Route commands:
    route add -net 192.168.10.1 netmask 255.255.255.0 gw 192.168.10.1
    route add -net 192.168.11.1 netmask 255.255.255.0 gw 192.168.11.1
    route add -net 192.168.12.1 netmask 255.255.255.0 gw 192.168.12.1

Site 1:
  eth1: internet ip: 2.2.3.4
  eth0: 192.168.10.1
  cipcb0 ip: 192.168.10.1 ptp: 192.168.1.1
  Route commands:
    route add -net 192.168.1.1 netmask 255.255.255.0 gw 192.168.1.1

Site 2:
  eth1: internet ip: 3.2.3.4
  eth0: 192.168.11.1
  cipcb0 ip: 192.168.11.1 ptp: 192.168.1.1
  Route commands:
    route add -net 192.168.1.1 netmask 255.255.255.0 gw 192.168.1.1

Site 3:
  eth1: internet ip: 4.2.3.4
  eth0: 192.168.12.1
  cipcb0 ip: 192.168.12.1 ptp: 192.168.1.1
  Route commands:
    route add -net 192.168.1.1 netmask 255.255.255.0 gw 192.168.1.1

Note that Site 1 can not get to Site 2 w/o adding a couple more routes to
Site 1&2, but they also don't need to.  The portland branch
(192.168.1.1) handles everything for centralized backup, SQL storage and
the like so the affiliate locations don't need to access eachother.  Also
note that portland (the central hub) has 4 interfaces whose ip is
192.168.1.1.  This is OK since the cipcbX interfaces are virtual, not
real.

When I first saw this, I thought it sounded hokey, but it works
great!  And it's much better than subnetting networks down to a.b.c.d/30
for the ptp link!

--Eric

On Fri, 17 Aug 2001, Peter van den Heuvel wrote:

> > Ahh!! You need your cipcb0 ip to be the same as an interface on your
> > system.  Try making your system's cipcb0 ip to 192.168.1.1.
> HUHHH???
> 
> I might not understand you coorectly, and thus, so might others.
> 
> Your cipcb0 interface is a (virtual) network interface.
> If you are a route (not a switch; as is mostly the case), you need a
> unique IP address on every interface and routes to move the data.
> Assigning one IP address to more than one interface is wrong.
> So, for a normal 2-peer cipe setup, you got 4 networks:
> 1 - private on A
> 2 - private on B
> 3 - point-to-point (cipe-to-cipe) so no need for these two to be in some
> sub-net
> 4 - the public net (mostly internet)
> 
> --
> Message sent by the cipe-l,AT,inka,DOT,de mailing list.
> Unsubscribe: mail majordomo,AT,inka,DOT,de, "unsubscribe cipe-l" in body
> Other commands available with "help" in body to the same address.
> CIPE info and list archive: 
><URL:http://sites.inka.de/~bigred/devel/cipe.html>
> 





<< | Thread Index | >> ]    [ << | Date Index | >> ]