Re: Notes from the battlefield
Doug Hughes (doug@eng.auburn.edu)
Sat, 16 Dec 95 12:01:06 CST
>
>1) discovery seems to get a bit confused in class B subnets -- so if I
>traceroute from, say, 147.72.56.1 to somewhere, the resulting diagram
>shows me connected to the 147.72 class B, and then that's conected to
>147.72.1, which is then connected to a router...when in reality I'm
>directly connected to the router via a slip line. this seems a little
>strange. Basically to get an accurate picture I'm forced to do the probe
>from -outside- of the network....
>
Yes, I have this problem too. Juergen and I discussed it and decided it
was because (in my case) the cisco router just returns the class B netmask
and not the subnet mask for the interface. One possible work-around that I've
thought about implementing, but never got around to, is to SNMP query
a device that returns a class-type (A,B,C) netmask to get the proper subnet
mask for the entering and exiting interface, if applicable.
>2) discovery seems to act inconsistently with routers with multiple (say,
>8) interfaces. For example, I can discover 2 or 3 networks and it
>correctly identifies them as all connected to the same router; discover
>others, and it creates a discrete object, even though it's just another
>interface of the original object. Is this a bug, or am I doing something
>wrong?
>
I've seen that too.. Know ideas there.. I just deleted the object and do
a trace to show the proper path, and that works okay.
>Likewise, it would be tres cool if you could limit -what- gets discovered
>(i.e. probe 192.221.244 and "only show ciscos").
>
This would definitely require a bit more smarts, perhaps via SNMP. It would
also slow down the discovery process I would think, but it would be an
interesting option...
Doug Hughes Engineering Network Services
doug@eng.auburn.edu Auburn University