[afnog] New IPv6 Address Block Allocated to the RIPE NCC

Brian Candler B.Candler at pobox.com
Wed Jan 4 00:25:10 EAT 2006


On Wed, Jan 04, 2006 at 07:37:45AM +1100, Geoff Huston wrote:
> addressing plus Network Address Translators. The legacy of this transition 
> is uncomfortable, with later adopters pointing to the somewhat liberal 
> address holdings of the early adopters and asking why they have to bear the 
                                                                  ^^^^^^^^^^^
> brunt of the cost and effort to achieve very high address utilization rates 
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> while the early adopters are still able to deploy relatively simple, but 
> somewhat more extravagant addressing schemes across their networks.

I think this will be much more true for IPv6, where there is a perception of
near-infinite address space availability, and therefore a much greater
resentment at being made to bear the cost and effort of high utilization.

> I suspect that this topic would be a useful AFRINIC policy discussion

Some key parts of the IPv6 addressing plan have ended up hardcoded into
stacks though - such as the 64 bit local token - and may be considered
out-of-scope for RIR policy. That's unfortunate I think, because it's one of
the fundamental problems.

If the techo-utopians involved in the initial design of IPv6 had simply
stuck to the tried and trusted mechanism of DHCP, there would be no problem.
You could just give everyone a /96 - which would be enough to run the whole
existing Internet in IPv4 - and let them get on with it. In fact, the choice
of 128 bits of IPv6 address would be considered extravagant, when 80 or 96
would have been plenty.

IIUC (and admittedly I don't have much operational experience of IPv6), an
IPv6 client still requires a router solicitation daemon, in which case it
might as well run DHCP. In any case, you probably want DHCP for picking up
other information such as DNS servers, time servers, TFTP boot servers and
so on.

Unfortunately, those who believe in IPv6 are so worried about its lack of
takeup even after all this time that they dare not make any changes to the
specs in case the changes prevent further takeup. Paradoxically, that might
be what makes the whole thing fail.

Regards,

Brian.



More information about the afnog mailing list