Re: IPv6 addresses really are scarce after all

[Tony Li <tli at cisco.com>]

> a clever router/switch could certainly do a lot without subnetting.


Indeed.  If you Google around a bit for "MAC table size", I think that
you'll find it hard to find an advertised size less than 1K entries,  
and that's
for a shrink-wrapped 8 port switch.

I'll believe that entry level boxes are smaller than that, but then  
once you
have more than a handful of systems, it would not be unreasonable
to invest a teeny amount in a better switch.

Just how big is your house/////TARDIS anyway?


> one of the areas in which I think the IPv4 design failed is that it
> didn't really follow the catenet model.  it was not possible to extend
> the network from any point.  and this is part of what led to NATs,
> because there really was a need to be able to do that.  IPv6 doesn't
> quite let you do that either, but having long addresses is probably
> "good enough" provided the prefixes given to supposed "end users" are
> still long enough to allow a couple of additional layers of network to
> be hung off of them.


The only way to do what you want is to effectively have a variable
length address.  While there were a few crazy advocates of this many
years ago, they were shouted down.

Tony Li
Lead Lunatic


_______________________________________________
Ietf mailing list
Ietf at ietf.org
https://www1.ietf.org/mailman/listinfo/ietf