Re: [sig-policy]Request to delay finalizing addressing proposals
At 23:11 28/11/2002 -0500, Thomas Narten wrote:
We've very recently become aware that APNIC has agreed to go forward
with three proposals relating to IPv6 address allocations:
- Documentation Prefix for IPv6
- Experimental Allocations (IPv6/IPv4/ASNs)
As author and co-author of these two APNIC policy proposals, I feel I ought
to comment.
We would like to request that APNIC temporarily delay considering
these policies finalized, in order for the IETF to have additional
time to study these issues and better coordinate with APNIC on topics
relating to the allocation of IPv6 address space for purposes other
than global unicast routing for LIRs and NIRs.
I don't understand this sentence. I thought that the IAB had delegated
responsibility for IPv6 address space allocation to the IANA, and that they
(the IAB) had requested that the IANA restrict allocations to 1/8th of the
current v6 space. Where did this requirement for global unicast routing for
LIRs and NIRs come from - do you have a reference anywhere?
We believe the IETF may
have relevant input to provide on these topics, but we have not
reviewed them because we were unaware of them.
So why couldn't the IETF have participated in the APNIC meeting in
September? The two policy proposals were on the APNIC website well before
the meeting happened, have been considered by the APNIC community,
discussed at length at the conference, and passed by the member meeting.
The experimental policy proposal has also been discussed at the RIPE
meeting during the LIR WG session (in what form I don't know as I couldn't
attend), and at the ARIN meeting. I've made the effort to ensure global
consensus with this, so I'm rather surprised that the IETF hasn't heard of it.
For your reference, you might like to consult
http://www.arin.net/policy/2002_2.html. The ARIN member meeting passed the
proposal, and it now goes to the ARIN board for approval.
RIPE is updating their existing documentation and that will go forward to
the next RIPE meeting for discussion and approval. You might like to
consult
http://www.ripe.net/ripe/draft-documents/experimental-resources.html which
was announced to the lir-wg mailing list on the 28th November, with closing
date for feedback on the 12th December. Or are the IETF unaware of this too?
Being unaware of what is going on in the operator community with this sort
of level of global discussion amongst the registries and their membership
at their public meetings seems quite worrying to me. I'm sorry I didn't
take my two proposals to the IETF, but quite frankly, they are operational
matters to do with address delegation and have nothing to do with the
establishment of standards.
With regards to the policy proposal:
Documentation Prefix for IPv6
http://www.apnic.net/meetings/14/sigs/policy/minutes.html#7
It should be noted that a similar proposal is currently a topic of
discussion within the IETF IPv6 Working Group.
Do you have a URL or other pointer? I try to keep up to date with I-Ds and
e-mail, but I may have missed some discussion.
In apparent contrast to the APNIC proposal, there is some
feeling that a larger allocation than a /32 is appropriate. We believe
it would be advantageous to have more general agreement on the
requirements for an IPv6 documentation allocation, prior to one
actually being made.
I suggested a /32 as that would be the path of least resistance to getting
acceptance from the global operator community. I'm happy to work with you
to put forward a modification to the policy proposal for consideration at
the next APNIC meeting which can suggest a larger allocation. We just need
to meet the deadlines for content suggestions for the APNIC policy meeting.
As with the /32, I'd like to get global consensus on what the revised size
should be - with the /32 size, both RIPE NCC and ARIN have indicated to me
that they will recommend the APNIC allocation if anyone approaches them
looking for an allocation for documentation purposes.
We also understand that there are also proposals on the topics of:
- Experimental Allocations (IPv6/IPv4/ASNs)
- Critical Infrastructure Assignments (IPv4/IPv6)
We have not had a chance to study these in detail and request that
their approval also be delayed, in order for the IETF community to
have a chance to study the issues more carefully. Additional time will
allow the IETF, IESG and IAB the opportunity to review these topics
and possibly provide a more formal statement, if they believe one is
needed after having reviewed the proposals.
Couldn't you have done this when the policy proposals were announced on the
APNIC website in August, rather than late on Friday night (Asia time) on
the weekend before the policy goes live?
I'd suggest that the best way to seek a review is to present an alternative
policy proposal at the APNIC meeting in February; and for the Experimental
Allocations policy you could also present an alternative proposal at the
upcoming RIPE and ARIN meetings? That's the way we all do it, and the
system does work.
Related documents that may be of interest:
ftp://ftp.ietf.org/internet-drafts/draft-narten-ipv6-iana-considerations-00.txt
(See Sections 2.4 and 3 in particular)
How is Section 2.4 relevant to this discussion? My two documents refer to
global unicast address space which is covered in Section 2.5. I see no
conflict here.
Section 3 simply summarises my policy proposal made to the APNIC/RIPE/ARIN
member meetings. You are more than welcome to refer to the APNIC
documentation /32 for this section.
best wishes!
philip
--
http://psg.com/~randy/draft-ymbk-sparse-v6-allocation-00.txt
http://psg.com/~randy/draft-ymbk-sparse-v6-allocation-00.html
Randy Bush,
IETF Operations Area Director (which includes the v6ops WG)
Thomas Narten,
IETF Internet Area Director (which includes the IPv6 WG)