2010-03 New Policy Proposal (Global Policy State in RIPE PDP)
PDP Number: 2010-03 Global Policy State in RIPE PDP Dear Colleagues A new RIPE Policy Proposal has been made and is now available for discussion. You can find the full proposal at: http://www.ripe.net/ripe/policies/proposals/2010-03.html We encourage you to review this proposal and send your comments to <address-policy-wg@ripe.net> before 18 May 2010. Regards Filiz Yilmaz Policy Development Manager RIPE NCC
Filiz Yilmaz wrote:
PDP Number: 2010-03 Global Policy State in RIPE PDP
The aim of this proposal seems to be worthwhile but I have a question and a note of caution. The question is that reading through the rationale I came across this statement "The benefits are that the RIPE community may speedily adopt a global policy proposal if it wishes, and that doing so will not delay the process of establishing consensus if further modifications are requested in other RIR communities." I can't see how the addition of an extra holding state can allow the policy to be adopted. Probably I'm confused, but surely this proposal just allows a proposal to be held in the new proposed state until all regions are in step. It won't allow the policy to be adopted any earlier? The note of caution is that there is no timeout on this state so we potentially have a policy held here for ever. Nigel
Hi Nigel, Good questions, thanks.
The question is that reading through the rationale I came across this statement
"The benefits are that the RIPE community may speedily adopt a global policy proposal if it wishes, and that doing so will not delay the process of establishing consensus if further modifications are requested in other RIR communities."
I can't see how the addition of an extra holding state can allow the policy to be adopted. Probably I'm confused, but surely this proposal just allows a proposal to be held in the new proposed state until all regions are in step. It won't allow the policy to be adopted any earlier?
I see what you mean. Let me characterise the problem like this: At this moment, if the RIPE community reaches consensus on a global policy, and it is adopted in our region before all other regions have adopted it, then there is still a risk that another region may propose a change that we want to adopt. This risk could, in principle, slow down the process of getting consensus in the RIPE region. The object of this proposal is to mitigate that risk, by allowing a proposal to be adopted in the RIPE region specifically, but ensuring that it can be revised if an alternative surfaces and reaches consensus. But a global policy of course can't be implemented until it has been adopted in all regions, and this proposal would not directly affect the speed at which that occurs.
The note of caution is that there is no timeout on this state so we potentially have a policy held here for ever.
Good point. There are two trapdoors though that can be "manually" triggered in the event that a policy seems to be stuck - one by the RIPE WG chairs, the other by the ASO AC. I have the feeling that adding a fixed timeout would probably not help the process, which can be lengthy. Does that sound reasonable? All the best, Dave -- Dave Wilson, Senior Network Engineer HEAnet Limited, Ireland's Education and Research Network 1st Floor, 5 George's Dock, IFSC, Dublin 1 Registered in Ireland, no 275301 tel: +353-1-660 9040 fax: +353-1-660 3666 web: http://www.heanet.ie/ H323 GDS:0035301101738 PGP: 1024D/C757ADA9
Dave Wilson wrote:
Hi Nigel,
Good questions, thanks.
I see what you mean. Let me characterise the problem like this:
At this moment, if the RIPE community reaches consensus on a global policy, and it is adopted in our region before all other regions have adopted it, then there is still a risk that another region may propose a change that we want to adopt. This risk could, in principle, slow down the process of getting consensus in the RIPE region.
The object of this proposal is to mitigate that risk, by allowing a proposal to be adopted in the RIPE region specifically, but ensuring that it can be revised if an alternative surfaces and reaches consensus. But a global policy of course can't be implemented until it has been adopted in all regions, and this proposal would not directly affect the speed at which that occurs.
So this allows a holding state, before the policy becomes instantiated in the RIPE region, from which it can be returned to a previous phase and allow changes (such as may have been made in another region), hence avoiding the startup delay? If so then I agree. Sounds like a fine idea.
Good point. There are two trapdoors though that can be "manually" triggered in the event that a policy seems to be stuck - one by the RIPE WG chairs, the other by the ASO AC. I have the feeling that adding a fixed timeout would probably not help the process, which can be lengthy. Does that sound reasonable?
Yes, it does. Nigel
So this allows a holding state, before the policy becomes instantiated in the RIPE region, from which it can be returned to a previous phase and allow changes (such as may have been made in another region), hence avoiding the startup delay?
Precisely this!
If so then I agree. Sounds like a fine idea.
All the best, Dave -- Dave Wilson, Senior Network Engineer HEAnet Limited, Ireland's Education and Research Network 1st Floor, 5 George's Dock, IFSC, Dublin 1 Registered in Ireland, no 275301 tel: +353-1-660 9040 fax: +353-1-660 3666 web: http://www.heanet.ie/ H323 GDS:0035301101738 PGP: 1024D/C757ADA9
This is perhaps more wordsmithing rather than a fundamental comment on the proposal, which sounds like A Good Thing. (1):
If the global policy proposal does not reach consensus or a substantial change is made on the proposal in one (or more) of the other RIR communities after the proposal was put in "Accepted pending consensus in other RIR communities" in RIPE, all the RIPE WG Chairs as a group will determine how to proceed. They can decide to withdraw the proposal or send it back to one of the previous phases of the RIPE PDP with or without a new version of the proposal.
(2):
If the global policy proposal fails to receive acceptance at the end of the global policy development process that is evaluated by the ASO Address Council (can be due to having substantial differences in the proposed text in different RIR communities or due to that the proposal failed to reach consensus in one of the RIR communities) then the proposal will be withdrawn automatically in RIPE too. The RIPE NCC will make the necessary announcements.
Is the second of these paragraphs needed? (Alternatively, is the first one needed?) They appear to cover similar events, so for clarity shouldn't it either be up to the WG chairs 'collective' or the output of the ASO AC? (As an aside, I'd remove 'all' from the first paragraph...) Cheers, Rob -- Rob Evans JANET(UK) Development Team Twitter: https://twitter.com/JANETDev/team Work tweets: https://twitter.com/internetplumber JANET(UK) is a trading name of The JNT Association, a company limited by guarantee which is registered in England under No. 2881024 and whose Registered Office is at Lumen House, Library Avenue, Harwell Science and Innovation Campus, Didcot, Oxfordshire. OX11 0SG
Hi Rob, Rob Evans wrote:
This is perhaps more wordsmithing rather than a fundamental comment on the proposal, which sounds like A Good Thing.
Thanks!
(1):
If the global policy proposal does not reach consensus or a substantial change is made on the proposal in one (or more) of the other RIR communities after the proposal was put in "Accepted pending consensus in other RIR communities" in RIPE, all the RIPE WG Chairs as a group will determine how to proceed. They can decide to withdraw the proposal or send it back to one of the previous phases of the RIPE PDP with or without a new version of the proposal.
(2):
If the global policy proposal fails to receive acceptance at the end of the global policy development process that is evaluated by the ASO Address Council (can be due to having substantial differences in the proposed text in different RIR communities or due to that the proposal failed to reach consensus in one of the RIR communities) then the proposal will be withdrawn automatically in RIPE too. The RIPE NCC will make the necessary announcements.
Is the second of these paragraphs needed? (Alternatively, is the first one needed?) They appear to cover similar events, so for clarity shouldn't it either be up to the WG chairs 'collective' or the output of the ASO AC?
I do see two subtly different circumstances here. The first is that a change has occurred during the process that gives the RIPE community reason to consider an amendment (or withdrawl) of the proposal. I think it's appropriate that this is triggered from inside the RIPE region. The second is something that I expect will only occur at the end of an unsuccessful process, and allows the RIPE NCC to tidy up once this has been established. We could certainly ask the WG chairs to rubber-stamp this, but I think it is a fairly clear state so I'm not sure there's much benefit to be gained from adding that step. All the best, Dave -- Dave Wilson, Senior Network Engineer HEAnet Limited, Ireland's Education and Research Network 1st Floor, 5 George's Dock, IFSC, Dublin 1 Registered in Ireland, no 275301 tel: +353-1-660 9040 fax: +353-1-660 3666 web: http://www.heanet.ie/ H323 GDS:0035301101738 PGP: 1024D/C757ADA9
Rob Evans wrote:
This is perhaps more wordsmithing rather than a fundamental comment on the proposal, which sounds like A Good Thing.
I also feel the proposal is A Good Thing, and I support it. As far as the comment is concerned, here are my 2 Euro cents:
(1):
If the global policy proposal does not reach consensus or a substantial change is made on the proposal in one (or more) of the other RIR communities after the proposal was put in "Accepted pending consensus in other RIR communities" in RIPE, all the RIPE WG Chairs as a group will determine how to proceed. They can decide to withdraw the proposal or send it back to one of the previous phases of the RIPE PDP with or without a new version of the proposal.
(2):
If the global policy proposal fails to receive acceptance at the end of the global policy development process that is evaluated by the ASO Address Council (can be due to having substantial differences in the proposed text in different RIR communities or due to that the proposal failed to reach consensus in one of the RIR communities) then the proposal will be withdrawn automatically in RIPE too. The RIPE NCC will make the necessary announcements.
Is the second of these paragraphs needed?
Well, in my opinion this is a very good question. I think the proposal 2010-03 assumes that the global policy proposal that has been accepted by the RIPE community and has been put in this new state, has been accepted only on condition it will eventually become a global policy and has no value if it remains a regional policy. In such a case, the automatic withdrawal in case of failure to become a global policy is the right thing to do. However, if the conditionally accepted policy has value as a regional policy too, then it may not be a good idea to automatically withdraw it. This is an argument in favour of deleting (2), as the RIPE WG Chairs could decide in accordance with (1) what to do in such a situation. This also raises the question whether it would be useful for the RIPE WG Chairs to be able to decide to keep a policy that is in this new state as a(n accepted) regional policy as it is (i.e without any further changes). In other words, the question is whether we want the RIPE WG Chairs to be able to move the policy from this new state to the Accepted state without having to consult the community, even if the policy would not become global. At present this possibility is not there, so either it becomes a _global_ policy as it is, or it goes back to the PDP (or it is withdrawn, of course). This could be solved by adding this possibility to (1). I personally think we should add this possibility.
(Alternatively, is the first one needed?)
Well, I definitely think it is important to have it there, as (2) deals only with the case when there is no chance to get a global policy in the end, while (1) deals with the case when some fine tuning could "save" the proposal. This fine tuning suggestion can come from the RIPE WG Chairs or from the community (when it is sent back in the same version).
They appear to cover similar events, so for clarity shouldn't it either be up to the WG chairs 'collective' or the output of the ASO AC?
I personally would be inclined to delete (2) and extend (1) as suggested above.
(As an aside, I'd remove 'all' from the first paragraph...)
I would also remove it, but I see no harm in having it there. I am sure the RIPE WG Chairs as a group will be able to reach consensus on what they recommend to do. Best regards, Janos
Hello all,
The objective of this policy proposal is to modify the PDP in a way specific to global policies. Global Policies (which, by definition, affect IANA), may not take effect until they have gained consensus in all five regions. In the case where RIPE adopts a global policy proposal but the proposal later changes in another region, this modification would make it explicit that RIPE may consider the changed text despite having already adopted an earlier text. Since RIPE 60, it's become clear to me that this is a more complex proposition than it looks. As I said at the meeting, the current Global Policy Development Process is very good at doing one thing: where a policy has consensus in all five regions, it is very good at determining this. By changing the RIPE PDP to try to make it easier to reach consensus where it does not yet exist, a number of complexities are introduced. In particular, on this list there have already been some questions about the cases enumerated where a GPP might not reach consensus. I think now that by changing the PDP in this way, we would have to enumerate many such cases, at various points of the PDP, and I fear we would not successfully cover them all. So while the goal is a laudable one, as I said at RIPE 60, I am not satisfied that this is an elegant way to solve it. The problem is not urgent, and we already have a way to deal with such a problem when it arises. So I do not think there is much to be gained by pushing forward in this way. On this basis, I would like to withdraw this proposal and, as suggested at RIPE 60, consider more generic alternatives. Best regards, Dave -- Dave Wilson, Senior Network Engineer HEAnet Limited, Ireland's Education and Research Network 1st Floor, 5 George's Dock, IFSC, Dublin 1 Registered in Ireland, no 275301 tel: +353-1-660 9040 fax: +353-1-660 3666 web: http://www.heanet.ie/ Calendar & PGP: http://people.heanet.ie/~davew/
participants (5)
-
Dave Wilson -
Filiz Yilmaz -
Janos Zsako -
Nigel Titley -
Rob Evans