Why is the IETF trying to micromanage ISP business models?
Is there a valid technical reason to allow for longer than /64 prefixes?
Regards,
-drc
On Sep 21, 2007, at 10:39 AM, Christian Huitema wrote:
This whole thread is based on an illusion.
The desire is to help small network operators subdivide the prefix
announced by their ISP.
In principle, if the ISP follows the current guidelines, there is
no issue: even the smallest network receives a /48, and deals with it.
In practice, some ISP are reluctant to follow these guidelines.
They profess concern about address space exhaustion, but we cannot
ignore the obvious marketing implications. If there is a two-tier
allocation, say /48 for enterprises and /64 for residential, then
ISP can use that in a two-tier tariff.
And here comes the illusion. If we observe that ISP only give a /64
to residential customers, then it is tempting to develop technology
so that /64 can be subdivided in multiple subnets, so that each
link gets some /(64+N) prefix. But has anyone stopped to ponder
what will happen next?
If ISP observe that a /64 prefix can be subdivided, and that a
single link can be addressed with /(64+N), what do you believe will
happen? How long will it take before we observe a different 3-tier
tariff, in which enterprises get /48, small businesses get /64, and
residential users get /(64+N)? And what will we do then? Develop
technology so that a /(64+N) prefix can be divided in /(64+N+P)
subnet prefixes?
-- Christian Huitema
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
