More Specifics of Classful Nets

Daniel Golding dgolding at
Tue Dec 7 03:16:06 UTC 1999


In fact, it would be a serious error to assume that a large provider
announces many routes. We advertise very few routes given our subscriber
base - few than my last ISP, which had only a fraction of the customers. A
scoring system is a bad idea because it penalizes and rewards based on
criteria other than network behavior and good net citizenship. If we could
buy and sell address space, with swamp blocks costing less than newer CIDR
blocks, your system would work, because it could encourage a mix that
would be advantagous (or not deleterious) to the net as a whole. Whoever,
the typical small Tier II has little ot no choice about the blocks it
gets. Whatever system that comes about in the regard must be fair for all
sizes of provider.

The solution is to allow all /24 and shorter prefixes in all address
space. Given current router memory and CPU capability, this is within our
means. I don't suggest de-aggregating these blocks, simply allowing
advertisement and propigation. 

Daniel L. Golding        *  Senior Network Engineer
Network Engineering      *  Mindspring Enterprises
dgolding at  * 

On Mon, 6 Dec 1999, Tim Wolfe wrote:

> On Mon, 6 Dec 1999, Steve E. Powell wrote:
> > What if the operator had the capability to accept routes based on
> > the ratio of good prefixes to bad prefixes received where a bad
> > prefix is defined as a more specific of a classful network or a
> So Provider A announces his ten /15 CIDR blocks to Provider B, aggregating
> huge numbers of IPs into those ten announcements.  He then wishes to
> announce 20 "bad" prefixes from customers of /21 or longer.  Under your
> plan, that doesn't fly.  Why is it that we assume that a "large" provider
> has to be announcing many routes?
> -- Tim
> --------------------------------------------------
> * Timothy M. Wolfe, Chief Network Engineer       *
> * ClipperNet Corporation / It's a wireless world *
> * tim at 800.338.2629 x 402	         *
> * Sufficient for today = Inadequate for tomorrow *
> --------------------------------------------------

More information about the NANOG mailing list