On 2007-07-05 03:55, Paul Vixie wrote:
[vixie]
my concern about aggregation is that if someone receives 8 /48's that are
aligned as a /45 then they could conceivably advertise it as a /45. i
hope that IANA will allocate nonaligned blocks, or perhaps give even
numbers to one RIR and odd numbers to the next, and that the RIRs and LIRs
will do likewise.
[carpenter]
There is a statistical sense in which ULA-G would be more *potentially*
aggregatable than ULA or ULA-C. If ULA-G prefixes are allocated on an
RIR/LIR basis, the chance that two logically adjacent prefixes happen to be
used by customers of the same local ISP will be finite, whereas it's
vanishingly small for ULA or ULA-C.
so perhaps we need to retain the randomness idea with instructions that IANA
and every RIR and LIR use it to ensure nonaggregatability?
When I think about it, this isn't such a big deal. If it's convenient
for an ISP to aggregate a few ULA-G prefixes on a local basis, for routing
within the ISP's domain, so what? The ULA prefix should still be filtered
at transit points.
I'm still not convinced about ULA-C/G but if this was an IESG ballot,
I'd be an 'abstain' at this point.
Brian
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
