[CUWiN-Dev] Re: IP Autoconfiguration scheme

[David Young]

On Mon, May 29, 2006 at 11:15:46PM +0200, Pablo Osuna wrote:
> Dear CUWIN community:
> 
> My name is Pablo Osuna. I am part of EHAS (www.ehas.org), a spanish
> non-profit institution (NGO), whose ultimate goal is improving public
> health assistance services in rural areas of Latin American Developing
> countries like Cuba, Peru and Colombia. We try to help them to develop
> the telecommunications infrastructure in rural establishments
> (particularly those that are the most isolated and lack phone line
> access).
> 
> I am the person in charge of developing a wireless mesh router with
> solar energy. I am working on that with an embedded Linux system. At the
> moment we have installed some networks in the countries mentioned before
> but we would like to improve the design, especially towards an
> autoconfiguration scheme.
> 
> The wireless router will be part of a mesh IP network and one of my main
> concerns is the IP Autoconfiguration. As you probably know there a lot
> of theoretical studios on this topic but we can not say the same in the
> implementation side.
> 
> I just saw your IP schema design. I copy it below:
> 
>      We will assign numbers to stations from 10.0/16. The last 16 bits
>         are the XOR of the first two octets of the MAC number, the second
>         two octets, and the third octets, where the bytes are taken in
>         "reading order."  That is, we produce numbers 10.0.A.B from the
>         MAC numbers. If the MAC produces A = 0, B = 0 or A = 255, B =
>         255, then A and B are assigned randomly.  The netmask is /16.
> 
>         (We compute numbers from the MAC to begin with because in the
>         common case, a station will boot with the same A and B every time,
>         which is useful for diagnostic purposes.)
> 
>         The host networks are assigned from 10/8. We assign to each
>         Ethernet interface, a network 10.A.B.0, where A and B are computed
>         as above from the Ethernet MAC number. If A = 0, we re-assign
>         it randomly. The netmask is /24.
> 
> Looking at this I would like to ask you: What happens if for example two
> nodes get the same IP Address after doing this procedure? I mean, the
> probability is very small, but that can happen. Have you predicted this?

If two nodes get the same IP address, then the network will break
in many predictable ways.  It will look to the routing protocol like
the two nodes are one node that has all the linkstates of both nodes.
At every node on the network, the routing protocol will choose just one
nexthop, the nexthop on the shortest path.  So the hop-wise more distant
node will ordinarily be the unreachable one.

Dave

-- 
David Young             OJC Technologies
dyoung at ojctech.com      Urbana, IL * (217) 278-3933