Keyboard Shortcuts
Thread View
j: Next unread messagek: Previous unread messagej a: Jump to all threadsj l: Jump to MailingList overview
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Dear SIG members,
The proposal, 'Additional criterion for final /8 allocations (and assignments)', has been sent to the Policy SIG for review. It will be presented at the Policy SIG at APNIC 31 in Hong Kong, 21-25 February 2011.
We invite you to review and comment on the proposal on the mailing list before the meeting.
The comment period on the mailing list before an APNIC meeting is an important part of the policy development process. We encourage you to express your views on the proposal:
- Do you support or oppose this proposal? - Does this proposal solve a problem you are experiencing? If so, tell the community about your situation. - Do you see any disadvantages in this proposal? - Is there anything in the proposal that is not clear? - What changes could be made to this proposal to make it more effective?
Information about this and other policy proposals is available from:
http://www.apnic.net/policy/proposals
Gaurab, Ching-Heng, and Terence
________________________________________________________________________
prop-089: Additional criterion for final /8 allocations (and assignments) ________________________________________________________________________
Author: David Woodhouse david@woodhou.se
Version: 1
Date: 11 January 2011
1. Introduction - ----------------
This proposal seeks to restrict the availability of IPv4 allocations and assignments from the final /8 to those members who are responsibly working to ease the transition period to IPv6 as we reach the final stages of IPv4 exhaustion.
2. Summary - -----------
APNIC shares with its members and their customers a collective responsibility to ensure manageable and scalable Internet growth[1].
It was once hoped that the transition to IPv6 would be completed long before the exhaustion of the IPv4 address space. The current 'crisis', if we can call it such, is can be attributed mostly to the fact that so many networks have been so slow to deploy IPv6.
Even today, many organisations are not yet ready to deploy IPv6, and are hoping to receive further allocations of IPv4 address space. Such a continued expansion of IPv4-only usage by those without a viable transition plan will help to drag out the transition period and exacerbate the issue for everyone.
A failure to deploy IPv6, even during the final stages of IPv4 exhaustion, demonstrates a clear failure to fulfill the above-mentioned responsibility.
3. Situation in other RIRs - ---------------------------
AfriNIC
The 'IPv4 Soft Landing Proposal' includes a provision that any network receiving IPv4 addresses during the proposed second part of the "exhaustion phase" defined in the proposal will also be delegated IPv6 addresses if they do not have any yet:
http://www.afrinic.net/docs/policies/AFPUB-2010-v4-005.htm
ARIN:
Policy Proposal 125, 'Efficient Utilization of IPv4 Requires Dual-Stack' was abandoned by the ARIN AC. The decision is currently being petitioned for re-inclusion of the proposal for discussion by the ARIN community:
http://lists.arin.net/pipermail/arin-ppml/2010-December/019054.html
RIPE:
Policy Proposal 2010-2, 'Allocations from the last /8' includes a provision that allocations shall only be made to LIRs if they have already received an IPv6 allocation.
http://www.ripe.net/ripe/policies/proposals/2010-02.html
There is no similar proposal in the LACNIC region.
4. Details - -----------
This is a proposal that amends the IPv4 allocation and assignment criteria to including the following additional criterion:
- To qualify for an allocation or assignment of IPv4, an organization should have a viable IPv6 deployment.
The meaning of what a "viable IPv6 deployment" is will depend on the type of organization is requesting the IPv4 addresses. A couple of examples are included below for clarity:
- Hosting service or organisation offering public-facing services
Each service the organisation offers on the newly-delegated IP addresses should be accessible by IPv6. It would not need to be on the same host, the same interface, or even in the same data center. If accessed by DNS, it should be accessed using the same hostname.
- ISP or network provider
The provider's default installation to new customers should include functional IPv6 connectivity, and IPv6 should be available to all end users on request within reasonable time/cost. As it isn't practical to require providers to instantly upgrade all existing equipment, this IPv6 requirement applies to new deployments which presumably use new equipment.
5. Pros/Cons - -------------
Advantages:
- This proposal will accelerate the adoption of IPv6 and help to reduce the transition period.
Even if the proposal fails completely in that aim, the proposed policy should still reduce the number of allocations to those organisations who would needlessly prolong the transition by deploying more IPv4-only services and networks at a time when all responsible organisations are fixing their existing legacy systems.
Disadvantages:
- Those organisation who had intended to further delay their inevitable adoption of IPv6, and exacerbate the problems by continuing to deploy IPv4-only services, will not be able to receive IPv4 addresses from the final /8. Effectively, the exhaustion will reach them just a little sooner than it reaches everybody else.
6. Effect on APNIC - -------------------
The principal effect would be accelerated adoption of IPv6 by APNIC members.
IPv4 exhaustion will affect APNIC members who have not yet taken steps to deal with it, before it affects those who have fulfilled their long-foreseen responsibility to adapt.
7. Effect on NIRs - ------------------
It is expected that NIRs would implement a matching policy for the final allocation phase.
8. References - --------------
[1] Section 6.3, " Collective responsibility" in "Policies for IPv4 address space management in the Asia Pacific region" http://www.apnic.net/policy/add-manage-policy#6.3
On 12/01/11 Wed, Jan 12, 23:58, Gaurab Raj Upadhaya wrote:
- Summary
I agree with David that we all have a responsibility to each other to keep the network working and endeavour to see dual stacked systems built. If we are ever to be in a position to move to an IPv6 only network then this can only happen after a critical dual stack mass is reached.
But I'm not convinced that this proposal is the way to do this.
- Situation in other RIRs
AfriNIC
The 'IPv4 Soft Landing Proposal' includes a provision that any network receiving IPv4 addresses during the proposed second part of the "exhaustion phase" defined in the proposal will also be delegated IPv6 addresses if they do not have any yet: http://www.afrinic.net/docs/policies/AFPUB-2010-v4-005.htmARIN:
Policy Proposal 125, 'Efficient Utilization of IPv4 Requires Dual-Stack' was abandoned by the ARIN AC. The decision is currently being petitioned for re-inclusion of the proposal for discussion by the ARIN community: http://lists.arin.net/pipermail/arin-ppml/2010-December/019054.htmlRIPE:
Policy Proposal 2010-2, 'Allocations from the last /8' includes a provision that allocations shall only be made to LIRs if they have already received an IPv6 allocation. http://www.ripe.net/ripe/policies/proposals/2010-02.htmlThere is no similar proposal in the LACNIC region.
Looking at these policies I'm encouraged that our policy of making IPv6 readily available to APNIC members is a good one. I'd like to think that as we move into the last /8 phase we'd mirror the AfriNIC approach. The IPv4 block being allocated at that point is for transition to IPv6 and so it makes sense to issue IPv6 space if none has already been issued.
If we want to make applicants think a little more about this then we could adopt the RIPE approach but given that our IPv6 allocation process is so simple it really wouldn't make much difference.
- Details
This is a proposal that amends the IPv4 allocation and assignment criteria to including the following additional criterion:
- To qualify for an allocation or assignment of IPv4, an organization should have a viable IPv6 deployment.
This is where my real concern lies. David gives a couple of examples of a "viable IPv6 deployment" which leaves us with the question of what other scenarios could be interpreted as "viable".
I'm concerned that unless we can come up with clear unambiguous definitions of what is acceptable here we will put the APNIC staff in the very difficult position of having to decide who can have IPv4 addresses and who can't based on ill-defined criteria. I suspect that disgruntled applicants could resort to legal challenges and that's not somewhere we want to go.
A number of people including me have expressed the view on this list and elsewhere that "IPv4 is over, finished...." and I think we should recognise this.
Our best option is not to keep adding patches and lashups to IPv4 policy in order to get IPv6 deployed - that will start to happen as the economic imperatives bite and not before.
Our IPv4 policy needs to remain clear and simple. During the last /8 each organisation will get a /22 (or whatever the size is at the time) for IPv6 transition purposes. I think it would make sense to allocate IPv6 space at the same time if they haven't got any.
(Apologies for broken threading; I'm replying to your message as I read it in the archives, and I can't see the Message-Id: so can't craft a correct reply. T'would be nice if the mailto: link in the archive page actually had a correct '?in-reply-to=foo&subject=bar' part...)
On Thu, 13 Jan 2011 at 10:21:06 +1300, Andy Linton wrote:
This is where my real concern lies. David gives a couple of examples of a "viable IPv6 deployment" which leaves us with the question of what other scenarios could be interpreted as "viable".
I'm concerned that unless we can come up with clear unambiguous definitions of what is acceptable here we will put the APNIC staff in the very difficult position of having to decide who can have IPv4 addresses and who can't based on ill-defined criteria. I suspect that disgruntled applicants could resort to legal challenges and that's not somewhere we want to go.
That was my initial concern, and my first draft of the proposal said that we would probably have to hash the details out on the policy list to come up with something that would be acceptable.
But I was advised that I was wrong — that we *could* actually rely on an element of common sense. With clarifying examples of what 'viable' would mean for the most common recipients of IP address allocations, I think that would be perfectly sufficient.
The question of what other scenarios could be interpreted as 'viable' is not one which should keep us awake at night. If a request comes in from an unusual type of member who doesn't fit any of the examples, and a 'common sense' answer is not apparent, we could even err on the side of permissiveness and allow the request. Such exceptions would be relatively few, and the policy would still have its intended effect, overall.
A number of people including me have expressed the view on this list and elsewhere that "IPv4 is over, finished...." and I think we should recognise this.
I agree with this observation. There's a reason I've been consistently referring to IPv4 as 'Legacy IP' for a number of years already. But the problem is that there are still people out there who *don't*, even though they should know better. Some examples: http://revk.www.me.uk/2011/02/what-hope-is-there.html http://revk.www.me.uk/2011/02/quote-of-day-most-cps-have-plenty-of.html
The point of the proposed policy is that for such people, IPv4 address space will have *already* run out. Because without a viable strategy for moving to IPv6, they will not be permitted to use any more of the few remaining Legacy IP addresses to prolong their fantasies.
Activity Summary
- 4674 days inactive
- 4674 days old
- sig-policy@lists.apnic.net
- 3 participants
- 2 comments