[c-nsp] Different behaviour for static route on different IOS?

[chooweikeong at pacific.net.sg]

Hi All,

Have to bring this up again, as i still cant conclude the exact behaviour 
of static route. From what i observe, this "problem" (suppress recursive 
static route) happens even the prefix and next-hop is not in the same 
supernet...


IOS (tm) 7200 Software (C7200-K4P-M), Version 12.0(22)S5, EARLY DEPLOYMENT 
RELEASE SOFTWARE (fc1)

interface Serial6/0
  ip address 192.154.x.x 255.255.255.252
  no ip directed-broadcast
  shutdown
  no cdp enable

ip classless
ip route 203.120.x.x 255.255.255.252 192.154.x.x

#sh ip in brief | in 192.154.x
Serial6/0           192.154.x.x   YES manual administratively down down

#sh ip route 203.120.x.x
% Subnet not in table

#sh ip route 192.154.x.x
Routing entry for 192.154.x.x/21, supernet
   Known via "ospf 10", distance 110, metric 20, type extern 2, forward metric 1


Hope you guys can help me out. Thanks.

Rgds,
Wei Keong


On Wed, 16 Feb 2005, Gert Doering wrote:

> Hi,
>
> On Wed, Feb 16, 2005 at 10:51:17AM +0100, Oliver Boehmer (oboehmer) wrote:
>>> Indeed, it's not "classful", but a more generic check "if the route
>>> prefix *and* the gateway IP are part of the *same* supernet route
>>> (/16, default route, ...) --> suppress recursive static route".
>>
>> coz it wouldn't make much sense, would it?
>>
>> you learn 192.168/16 via POS0/0, and you enter a static route for
>> 192.168.20/24 pointing to a nexthop within 192.168/16, all traffic to
>> 192.168/16 (including 192.168.20/24) would still go out POS0/0.
>
> Well.  There is something to be said for this...
>
>> because it adds routing information.. I don't really find this
>> surprising..
>
> ... but usually the box doesn't check "does it add any useful information
> if I put this route into service" either.
>
> Like in:
>
>  ip route 192.168.0.0 255.255.0.0 pos0/0
>  ip route 192.168.20.0 255.255.255.0 pos0/0
>
> -> you'll see *both* routes in the RIB and FIB, even if the second route
> "doesn't add routing information" either...
>
> But whatsoever - I'm not trying to criticise (today :) ), I was just
> explaining my surprise, and now I have understood what's going on.
>
> gert
> --
> USENET is *not* the non-clickable part of WWW!
>                                                           //www.muc.de/~gert/
> Gert Doering - Munich, Germany                             gert at greenie.muc.de
> fax: +49-89-35655025                        gert at net.informatik.tu-muenchen.de
>