[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
draft-despres-v6ops-6rd-ipv6-rapid-deployment
- To: IPv6 Operations <v6ops@ops.ietf.org>
- Subject: draft-despres-v6ops-6rd-ipv6-rapid-deployment
- From: james woodyatt <jhw@apple.com>
- Date: Tue, 11 Mar 2008 14:44:35 -0400
everyone--
I suppose I should offer comments on this draft, too, since it
envisions things happening in residential gateway CPE that do not
currently happen today.
It seems to me that 2002::/16, the 6to4 prefix, can be regarded as a
special case of 6rd ISP prefix allocated by IANA and unassigned to any
particular organization. What 6RD does is allow each ISP to define
their own 6to4-like routing domain, provided that CPE functions are
properly loaded with the 6rd ISP prefix to use as distinguished from
the 2002::/16 prefix assigned by IANA.
The draft defines a new DHCPv4 option used by 6RD ISP to configure 6RD-
enabled CPE, i.e. hosts or dual-stack IPv4/IPv6 routers with 6RD-
tunneling functions. I very much doubt that getting vendors of
consumer retail CPE, e.g. host operating systems, consumer routers,
etc., to adopt this DHCP client feature will be easy. For that
reason, I don't expect 6RD to be deployable except with provider-
provisioned CPE.
Finally, the section where it's noted that RFC 1918 addresses can be
used to construct 6RD site addresses is interesting, but it implies
that those ISP's which are large enough to provide service to more
IPv4 endpoints than will fit inside the RFC 1918 portion of the
addressing realm will need to use more than one 6RD ISP prefix to
disambiguate the realms. Does this conflict with the statement on
page 6, "For this, a simple approach consists in each 6rd ISP to chose
one and only one of its IPv6 unicast prefixes as the '6rd ISP prefix'
which appears at the start of 6rd site addresses addresses, and to
have this prefix known by 6rd CPEs." It seems to me that such ISP's
will be required to operate more than one 6RD relay and ensure that
interior routes are plumbed properly for traffic to flow between
them. To do this, it seems that such ISP's will need more than one
RIR-allocated prefix. Is that a problem?
Also, multicast?
--
james woodyatt <jhw@apple.com>
member of technical staff, communications engineering