Routing, Forwarding, and the Legacy Internet

Chrisy Luke chrisy at flix.net
Sun Sep 13 22:11:05 UTC 1998


Jim Fleming wrote (on Sep 13):
>                         Let's also imagine that your
> forwarder (not router) could have up to 256 interfaces.
> With 24 bits and a 1 byte wide table, you would need
> 16,777,216 bytes to describe all possible /24s and
> the interface a packet should be forwarded on.

This is a very naieve way of looking at how a forwarding table should
be stored, particularly since it doesn't address many issues related
to internal network topology, imbalance of routes across interfaces,
statistics gathered about use of each entry - including it's age, where
it learnt it from, other sources (and next-hops) for the same destination,
metrics, load-balancing in the case of multipath capable routers etc etc,
not to mention the burden of physically transferring this data to the 
router in the first place.

Even in your simplistic view, you would have to transfer 16meg of data
over a network to get a complete table. Large portions of the world
still use T/E-1's etc on BGP speaking lines. 16mbytes over that will take
a looong time.  And then you have all the BGP metrics to consider - AS path,
MEDs and so forth.

> is not a lot of memory in this day and age. Also, once
> the table was filled in, the lookup would be very fast.

Only because you assume "all networks are equal". They aren't.

Chris.
-- 
== chris at easynet.net, chrisy at flix.net, chrisy at flirble.org
== Systems Manager for Easynet, part of Easynet Group PLC.



More information about the NANOG mailing list