Re: [address-policy-wg] 2015-03 New Draft Document and Impact Analysis Published (Assessment Criteria for IPv6 Initial Allocation Size)
Hi to all +1 I fully support this proposal. I believe it is an important step to allow for useful deployment of IPv6 taking advantage of the advanced and extended IPv6 address architecture. Responding to the request for feedback on the "RIPE understanding" section here are some thoughts and comments: - The overall tone of the "understanding" section still sounds more like "our main goal is to preserve addresses" instead of "let's make generous use oft he extended address architecture and available space" to me. It is not easy to step out of IPv4 thinking. While we all know in our minds that we need to do this, implementing it is much more difficult. (I am for instance referring to this sentence, which appears more than once: " The RIPE NCC will ask for additional documentation to justify why a less address consuming hierarchy or topology can not be implemented."). There is a widely adopted rule that all address conservation mechanisms should be removed from IPv6 address plans. - There is one point where I have a problem to understand, maybe I am misinterpreting the statement. It says " The RIPE NCC will consider longevity reasonable for a similar timeframe for which past growth was documented." - So my question is: how does this accomodate startups? We are moving into the age of the Internet of Things. We expect new technologies and services to spring up, not foreseeable. This includes new business models and opportunities for new companies that have no history. - The policy for subsequent allocations will have to be updated accordingly if this policy is accepted. - Quote: "If this network topology is justified, the RIPE NCC will consider up to one extra bit per hierarchical level or geographical segment as reasonable, on top of the documented need for that part of the network." Comment: how about "generally up to one bit" - leave room for exceptions. - The rules seem quite complicated and a bit hard to really understand and correctly apply to me. I wonder how easy it will be to assess requests based on this. As mentioned already, I believe this proposal is a good step. When it comes to how to apply it, I feel there is a lot of "strictness" in the wording. It is always nice to have clear and strict rules, everyone can easily apply them. But reality is different. There is no good rule without exceptions and I hope that this policy will be applied with common sense and adjusted to reality. And my personal viewpoint is, that we should not restrict IPv6 address plans from the beginning. Our main target should be to finally DEPLOY it as broadly as possible and have ease of operation as a main goal. That is what the address architecture provides. By today we have given out the equivalent of 171'000 /32 of the currently defined global unicast space (2000::/3) (according to Bgpexpert<http://www.bgpexpert.com/addrspace-ipv6.php>) and this equals 0.032%. Hey that is 171'000 times more than the whole current Internet and we are at 0.032%! I think we should apply a generous allocation model that makes it easy to deploy, operate, secure..... ...and if we get to the point where we feel we have been too generous, we can still adjust policies and become more careful. Once we have used 2000::/3 we still have 7 more chances to do better. We could even define a re-assessment once we are 30% or 50% into the 2000::/3. My two cents, greetings from the summer heat Silvia Hagen Chair Swiss IPv6 Council -----Ursprüngliche Nachricht----- Von: address-policy-wg [mailto:address-policy-wg-bounces@ripe.net] Im Auftrag von Marco Schmidt Gesendet: Donnerstag, 9. Juli 2015 14:20 An: policy-announce@ripe.net<mailto:policy-announce@ripe.net> Cc: address-policy-wg@ripe.net<mailto:address-policy-wg@ripe.net> Betreff: [address-policy-wg] 2015-03 New Draft Document and Impact Analysis Published (Assessment Criteria for IPv6 Initial Allocation Size) Dear colleagues, The draft document for version 2.0 of the policy proposal 2015-03, "Assessment Criteria for IPv6 Initial Allocation Size", has now been published, along with an impact analysis conducted by the RIPE NCC. The proposal aims to expand the criteria for evaluating initial IPv6 allocations larger than a /29. The RIPE NCC would consider additional aspects beyond only the number of existing users and extent of the organisation's infrastructure. Some of the differences from version 1.0 include: - Introduction of new assessment criteria used to evaluate IPv6 allocations larger than a /29 - Related wording adjustments in the summary and rationale of the proposal You can find the full proposal and the impact analysis at: https://www.ripe.net/participate/policies/proposals/2015-03 and the draft document at: https://www.ripe.net/participate/policies/proposals/2015-03/draft We encourage you to read the draft document text and send any comments to address-policy-wg@ripe.net<mailto:address-policy-wg@ripe.net> before 7 August 2015. Regards Marco Schmidt Policy Development Officer RIPE NCC Sunny Connection AG CH-8124 Maur +41 (0)44 887 62 10 www.sunny.ch<http://www.sunny.ch/> http://twitter.com/sunny_shagen ****************************************** IPv6 Business Conference, June 18, 2015, Zurich<http://www.ipv6conference.ch/> International Top Speakers, Unique Program Great Networking - watch out for the 2016 Conference ****************************************** Switzerland is in the Top 5 in IPv6 User Adoption We reached 21% in June 2015! *********************************** Our website is dual-stack - how about yours? ***********************************
Hi, On Fri, Jul 24, 2015 at 11:31:54AM +0000, Silvia Hagen wrote:
There is a widely adopted rule that all address conservation mechanisms should be removed from IPv6 address plans.
You can't do that on a RIR level - if the IPRAs were to hand out a /16 for everyone that comes with a nice diagram, we'd actually run out of IPv6 soon. Of course a /16 is excaggerating a bit - but I have seen my share of network plans made totally without understanding for bits, hierarchy or actual *networking*, resulting in "oh, for these 500 sites, we definitely need a /24!" (and "oh, for all the electronic passports for 100 million citizens, we must have a /19!") - and thus it is good practice to have someone more experienced in addressing review the plan and see whether it makes sense. (Just to point out the obvious - from the early days of /35s I have been fighting for more liberal IPv6 allocation policies, but it still needs to be done with a solid technical understanding, and not with "I like large numbers, so get me a /15 please!" - this is the balance we need to find, or otherwise we'll find us faster than expected in the "oops, fp 001 is gone!" land) Gert Doering -- APWG chair -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
Hi Gert Sure, I fully agree with what you are saying, that is actually what I meant with "use common sense". So we add to that "and with the necessary technical understanding". The reason that I made the statement from this perspective is that in my consultings I have seen a lot more oft he restricted thinking (like when a global organization says: "we got a /48 and I guess we will find a way to live with that, it is more than we ever had".... :-) So let's go for balance :-) Silvia -----Ursprüngliche Nachricht----- Von: Gert Doering [mailto:gert@space.net] Gesendet: Freitag, 24. Juli 2015 13:51 An: Silvia Hagen Cc: address-policy-wg@ripe.net Betreff: Re: [address-policy-wg] 2015-03 New Draft Document and Impact Analysis Published (Assessment Criteria for IPv6 Initial Allocation Size) Hi, On Fri, Jul 24, 2015 at 11:31:54AM +0000, Silvia Hagen wrote:
There is a widely adopted rule that all address conservation mechanisms should be removed from IPv6 address plans.
You can't do that on a RIR level - if the IPRAs were to hand out a /16 for everyone that comes with a nice diagram, we'd actually run out of IPv6 soon. Of course a /16 is excaggerating a bit - but I have seen my share of network plans made totally without understanding for bits, hierarchy or actual *networking*, resulting in "oh, for these 500 sites, we definitely need a /24!" (and "oh, for all the electronic passports for 100 million citizens, we must have a /19!") - and thus it is good practice to have someone more experienced in addressing review the plan and see whether it makes sense. (Just to point out the obvious - from the early days of /35s I have been fighting for more liberal IPv6 allocation policies, but it still needs to be done with a solid technical understanding, and not with "I like large numbers, so get me a /15 please!" - this is the balance we need to find, or otherwise we'll find us faster than expected in the "oops, fp 001 is gone!" land) Gert Doering -- APWG chair -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
Hi, On Fri, Jul 24, 2015 at 12:02:31PM +0000, Silvia Hagen wrote:
So let's go for balance :-)
All for it! Plus "good documentation and good understanding of networking" Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
On Fri, Jul 24, 2015 at 12:02:31PM +0000, Silvia Hagen wrote:
So let's go for balance :-)
I agree. I think a sensible balance may be that allocations >/29 are reviewed (as they are now, AIUI) by the IPRA managers and/or the Board. There is a danger, in my opinion, that the IPv6 allocation/assignment process is infested with "IPv4 thinking" which will result in SPs employing "workarounds" of the sort that made IPv4 such a pain to deal with. That said, I've done a bit of v6 IPAM recently and the numbers one deals with are staggering. ;p rgds, Sascha Luck
Hi Gert,
I have seen my share of network plans made totally without understanding for bits, hierarchy or actual *networking*, resulting in "oh, for these 500 sites, we definitely need a /24!" (and "oh, for all the electronic passports for 100 million citizens, we must have a /19!") - and thus it is good practice to have someone more experienced in addressing review the plan and see whether it makes sense.
I think that is the general point that Silvia is making... It is very difficult, if not impossible, to come up with a one-size-fits all approach to assessing requests for address space, particularly when trying to cater for organisations that don't quite fit the usual mould. Are approaches such as 'up to one extra bit per hierarchical level or geographical segment' compatible with this premise and are they even necessary? I know from personal experience of assessing hundreds of requests for IPv4 address ranges over the years within my own organisation that there is no substitute for experience when it comes to performing the task effectively. Whilst rules of thumb are useful I think that attempts to 'proceduralise' the task with more specifics can end up being unhelpful and, in any event, are not necessary prerequisites for consistency. In my view what is more important is general oversight and capturing of experience garnered over multiple requests and it is noted from the IA that >/29 requests will continue to follow the escalated evaluation process which ought to help provide this. To be clear about where I stand; I am still satisfied that the revised policy and its proposed implementation will meet the needs of the UK MOD however I would nevertheless support a more 'liberal' approach to the consideration of the varying requirements of other organisations who will undoubtedly have different - yet equally valid - priorities and needs. Regards, Mathew
Gert, Apologies I missed this bit in my first response:
(Just to point out the obvious - from the early days of /35s I have been fighting for more liberal IPv6 allocation policies, but it still needs to be done with a solid technical understanding, and not with "I like large numbers, so get me a /15 please!" - this is the balance we need to find, or otherwise we'll find us faster than expected in the "oops, fp 001 is gone!" land)
I fully agree with this but would ask that you are careful not to inadvertently imply that all large requests are necessarily a result of technical ignorance and/or disregard for what is still a finite shared resource. Some will be I am sure but there will be many that are the result of considerable effort being put into identifying requirements and establishing a sensible and robust addressing strategy to satisfy them. I think we should be careful that we don't focus too much on constraining the former but end up having an even bigger negative impact on the latter. As you say; it really is all about balance. Mathew
Hi Silvia, Thanks for your feedback on our understanding of the policy. Regarding your point about start-ups - it can be difficult to make a fair evaluation on what is often a vague estimate of potential growth, and so the tendency here is to be somewhat cautious. On the other hand, we feel that our approach under the existing policy has been working well, and we have never had any issues or complaints from start-ups in the past. I would also like to respond to the point that both you and Matthew made about the RIPE NCC taking a more liberal approach in our interpretation. There is a balance to be struck here, between allowing for corner cases, and the requirements being clearly listed and adhered to. Our understanding in the impact analysis is based both on our previous experience with IPv6 requests and our interpretation of the policy text. If the community would like us to take a more liberal approach, we will need some additional guidelines on how to evaluate the requests in the proposed policy text. Kind regards Marco Schmidt Policy Development Officer RIPE NCC On 24/07/15 13:31, Silvia Hagen wrote:
Hi to all
+1
I fully support this proposal. I believe it is an important step to allow for useful deployment of IPv6 taking advantage of the advanced and extended IPv6 address architecture.
Responding to the request for feedback on the "RIPE understanding" section here are some thoughts and comments:
-The overall tone of the "understanding" section still sounds more like "our main goal is to preserve addresses" instead of "let's make generous use oft he extended address architecture and available space" to me.
It is not easy to step out of IPv4 thinking. While we all know in our minds that we need to do this, implementing it is much more difficult. (I am for instance referring to this sentence, which appears more than once: " The RIPE NCC will ask for additional documentation to justify why a less address consuming hierarchy or topology can not be implemented.").
There is a widely adopted rule that all address conservation mechanisms should be removed from IPv6 address plans.
-There is one point where I have a problem to understand, maybe I am misinterpreting the statement. It says " The RIPE NCC will consider longevity reasonable for a similar timeframe for which past growth was documented." - So my question is: how does this accomodate startups? We are moving into the age of the Internet of Things. We expect new technologies and services to spring up, not foreseeable. This includes new business models and opportunities for new companies that have no history.
-The policy for subsequent allocations will have to be updated accordingly if this policy is accepted.
-Quote: "If this network topology is justified, the RIPE NCC will consider up to one extra bit per hierarchical level or geographical segment as reasonable, on top of the documented need for that part of the network." Comment: how about "generally up to one bit" - leave room for exceptions.
-The rules seem quite complicated and a bit hard to really understand and correctly apply to me. I wonder how easy it will be to assess requests based on this.
As mentioned already, I believe this proposal is a good step. When it comes to how to apply it, I feel there is a lot of "strictness" in the wording. It is always nice to have clear and strict rules, everyone can easily apply them. But reality is different. There is no good rule without exceptions and I hope that this policy will be applied with common sense and adjusted to reality.
And my personal viewpoint is, that we should not restrict IPv6 address plans from the beginning. Our main target should be to finally DEPLOY it as broadly as possible and have ease of operation as a main goal. That is what the address architecture provides.
By today we have given out the equivalent of 171'000 /32 of the currently defined global unicast space (2000::/3) (according to Bgpexpert <http://www.bgpexpert.com/addrspace-ipv6.php>) and this equals 0.032%. Hey that is 171'000 times more than the whole current Internet and we are at 0.032%! I think we should apply a generous allocation model that makes it easy to deploy, operate, secure…..
…and if we get to the point where we feel we have been too generous, we can still adjust policies and become more careful. Once we have used 2000::/3 we still have 7 more chances to do better. We could even define a re-assessment once we are 30% or 50% into the 2000::/3.
My two cents, greetings from the summer heat
Silvia Hagen
Chair Swiss IPv6 Council
-----Ursprüngliche Nachricht----- Von: address-policy-wg [mailto:address-policy-wg-bounces@ripe.net] Im Auftrag von Marco Schmidt Gesendet: Donnerstag, 9. Juli 2015 14:20 An: policy-announce@ripe.net <mailto:policy-announce@ripe.net> Cc: address-policy-wg@ripe.net <mailto:address-policy-wg@ripe.net> Betreff: [address-policy-wg] 2015-03 New Draft Document and Impact Analysis Published (Assessment Criteria for IPv6 Initial Allocation Size)
Dear colleagues,
The draft document for version 2.0 of the policy proposal 2015-03, "Assessment Criteria for IPv6 Initial Allocation Size", has now been published, along with an impact analysis conducted by the RIPE NCC.
The proposal aims to expand the criteria for evaluating initial
IPv6 allocations larger than a /29. The RIPE NCC would consider additional aspects beyond only the number of existing users and extent of the organisation's infrastructure.
Some of the differences from version 1.0 include:
- Introduction of new assessment criteria used to evaluate IPv6 allocations larger than a /29
- Related wording adjustments in the summary and rationale of the proposal
You can find the full proposal and the impact analysis at:
https://www.ripe.net/participate/policies/proposals/2015-03
and the draft document at:
https://www.ripe.net/participate/policies/proposals/2015-03/draft
We encourage you to read the draft document text and send any comments to address-policy-wg@ripe.net <mailto:address-policy-wg@ripe.net> before 7 August 2015.
Regards
Marco Schmidt
Policy Development Officer
RIPE NCC
Sunny Connection AG
CH-8124 Maur
+41 (0)44 887 62 10
www.sunny.ch <http://www.sunny.ch/>
http://twitter.com/sunny_shagen
******************************************
IPv6 Business Conference, June 18, 2015, Zurich <http://www.ipv6conference.ch/>__
International Top Speakers, Unique Program
Great Networking - *watch out for the 2016 Conference*
******************************************
Switzerland is in the Top 5 in IPv6 User Adoption
We reached 21% in June 2015!
***********************************
Our website is dual-stack – how about yours?
***********************************
Hi Marco,
I would also like to respond to the point that both you and Matthew made about the RIPE NCC taking a more liberal approach in our interpretation. There is a balance to be struck here, between allowing for corner cases, and the requirements being clearly listed and adhered to. Our understanding in the impact analysis is based both on our previous experience with IPv6 requests and our interpretation of the policy text. If the community would like us to take a more liberal approach, we will need some additional guidelines on how to evaluate the requests in the proposed policy text.
Perhaps this is where the difficulty lies... A liberal approach generally presupposes the absence of rules and/or precise definition and so trying to put specifics into policy to promote such an approach may be counter-productive. On that basis I think I'd prefer not to see the proposed policy text containing even more detail than it does now. I guess this is where the Impact Analysis, and its publication, plays such a key role - it helps form a common understanding between RIPE NCC and the community as to how the policy can/will be implemented without requiring the policy text to be so specific. Taking this a step further it can also allow for a pragmatic approach to be taken where necessary and appropriate i.e. it captures the spirit of the law as opposed to the letter. Mathew
Hi Marco Startups: If you have an approach that works well, cool! A startup can have big plans and have success or fail or have to scale down substantially, there is always a risk. But we never know in advance. If you have a policy that works well, thats good. Probably the point will be to "reserve" enough space in case they turn into the next Google or FB (in size) :) Liberal approach: To me, liberal means no too strict rules. You can define as many rules as you want, you will never be able to catch and cover all possible corner cases with it. We (community) cannot give absolution to any decision you make by writing guidelines that do so either. My personal stance is: I trust you RIPE people to understand the issues, the market, the IPv6 addressing architecture and requirements well enough to make informed, balanced and fair decisions that support deployment of IPv6 when assessing requests. I think we are all good to go as it is. Silvia Von: Marco Schmidt [mailto:mschmidt@ripe.net] Gesendet: Montag, 27. Juli 2015 16:01 An: Silvia Hagen; address-policy-wg@ripe.net Betreff: Re: [address-policy-wg] 2015-03 New Draft Document and Impact Analysis Published (Assessment Criteria for IPv6 Initial Allocation Size) Hi Silvia, Thanks for your feedback on our understanding of the policy. Regarding your point about start-ups - it can be difficult to make a fair evaluation on what is often a vague estimate of potential growth, and so the tendency here is to be somewhat cautious. On the other hand, we feel that our approach under the existing policy has been working well, and we have never had any issues or complaints from start-ups in the past. I would also like to respond to the point that both you and Matthew made about the RIPE NCC taking a more liberal approach in our interpretation. There is a balance to be struck here, between allowing for corner cases, and the requirements being clearly listed and adhered to. Our understanding in the impact analysis is based both on our previous experience with IPv6 requests and our interpretation of the policy text. If the community would like us to take a more liberal approach, we will need some additional guidelines on how to evaluate the requests in the proposed policy text. Kind regards Marco Schmidt Policy Development Officer RIPE NCC On 24/07/15 13:31, Silvia Hagen wrote: Hi to all +1 I fully support this proposal. I believe it is an important step to allow for useful deployment of IPv6 taking advantage of the advanced and extended IPv6 address architecture. Responding to the request for feedback on the "RIPE understanding" section here are some thoughts and comments: - The overall tone of the "understanding" section still sounds more like "our main goal is to preserve addresses" instead of "let's make generous use oft he extended address architecture and available space" to me. It is not easy to step out of IPv4 thinking. While we all know in our minds that we need to do this, implementing it is much more difficult. (I am for instance referring to this sentence, which appears more than once: " The RIPE NCC will ask for additional documentation to justify why a less address consuming hierarchy or topology can not be implemented."). There is a widely adopted rule that all address conservation mechanisms should be removed from IPv6 address plans. - There is one point where I have a problem to understand, maybe I am misinterpreting the statement. It says " The RIPE NCC will consider longevity reasonable for a similar timeframe for which past growth was documented." - So my question is: how does this accomodate startups? We are moving into the age of the Internet of Things. We expect new technologies and services to spring up, not foreseeable. This includes new business models and opportunities for new companies that have no history. - The policy for subsequent allocations will have to be updated accordingly if this policy is accepted. - Quote: "If this network topology is justified, the RIPE NCC will consider up to one extra bit per hierarchical level or geographical segment as reasonable, on top of the documented need for that part of the network." Comment: how about "generally up to one bit" - leave room for exceptions. - The rules seem quite complicated and a bit hard to really understand and correctly apply to me. I wonder how easy it will be to assess requests based on this. As mentioned already, I believe this proposal is a good step. When it comes to how to apply it, I feel there is a lot of "strictness" in the wording. It is always nice to have clear and strict rules, everyone can easily apply them. But reality is different. There is no good rule without exceptions and I hope that this policy will be applied with common sense and adjusted to reality. And my personal viewpoint is, that we should not restrict IPv6 address plans from the beginning. Our main target should be to finally DEPLOY it as broadly as possible and have ease of operation as a main goal. That is what the address architecture provides. By today we have given out the equivalent of 171'000 /32 of the currently defined global unicast space (2000::/3) (according to Bgpexpert<http://www.bgpexpert.com/addrspace-ipv6.php>) and this equals 0.032%. Hey that is 171'000 times more than the whole current Internet and we are at 0.032%! I think we should apply a generous allocation model that makes it easy to deploy, operate, secure..... ...and if we get to the point where we feel we have been too generous, we can still adjust policies and become more careful. Once we have used 2000::/3 we still have 7 more chances to do better. We could even define a re-assessment once we are 30% or 50% into the 2000::/3. My two cents, greetings from the summer heat Silvia Hagen Chair Swiss IPv6 Council -----Ursprüngliche Nachricht----- Von: address-policy-wg [mailto:address-policy-wg-bounces@ripe.net] Im Auftrag von Marco Schmidt Gesendet: Donnerstag, 9. Juli 2015 14:20 An: policy-announce@ripe.net<mailto:policy-announce@ripe.net> Cc: address-policy-wg@ripe.net<mailto:address-policy-wg@ripe.net> Betreff: [address-policy-wg] 2015-03 New Draft Document and Impact Analysis Published (Assessment Criteria for IPv6 Initial Allocation Size) Dear colleagues, The draft document for version 2.0 of the policy proposal 2015-03, "Assessment Criteria for IPv6 Initial Allocation Size", has now been published, along with an impact analysis conducted by the RIPE NCC. The proposal aims to expand the criteria for evaluating initial IPv6 allocations larger than a /29. The RIPE NCC would consider additional aspects beyond only the number of existing users and extent of the organisation's infrastructure. Some of the differences from version 1.0 include: - Introduction of new assessment criteria used to evaluate IPv6 allocations larger than a /29 - Related wording adjustments in the summary and rationale of the proposal You can find the full proposal and the impact analysis at: https://www.ripe.net/participate/policies/proposals/2015-03 and the draft document at: https://www.ripe.net/participate/policies/proposals/2015-03/draft We encourage you to read the draft document text and send any comments to address-policy-wg@ripe.net<mailto:address-policy-wg@ripe.net> before 7 August 2015. Regards Marco Schmidt Policy Development Officer RIPE NCC Sunny Connection AG CH-8124 Maur +41 (0)44 887 62 10 www.sunny.ch<http://www.sunny.ch/> http://twitter.com/sunny_shagen ****************************************** IPv6 Business Conference, June 18, 2015, Zurich<http://www.ipv6conference.ch/> International Top Speakers, Unique Program Great Networking - watch out for the 2016 Conference ****************************************** Switzerland is in the Top 5 in IPv6 User Adoption We reached 21% in June 2015! *********************************** Our website is dual-stack - how about yours? ***********************************
participants (5)
-
Gert Doering -
Marco Schmidt -
Mathew Newton -
Sascha Luck [ml] -
Silvia Hagen