Orbit

I suppose the question we don't know the answer to is "how many organisations simply don't bother applying for IPv6 space because they believe they MUST deploy to 200 sites within 2 years".

Note my wording, as that is the implication from the problem that the authors of the proposal want to solve.

Thanks for clarifying. This is indeed is the issue we are trying to solve.(i.e,to prevent people from misunderstanding the criteria and be reserved from make requests despite their needs.)

We found out that out of 60 members we visited last year, more than half of small to medium LIRs haven't bothered to apply for allocations in the first place thinking they can't meet "200 assigments" criteria. This includes those who wish to start deployment as a test case first and not 100% sure to start as business, so the figure could be lower if you seek for those solid service plan. It still did help us see that there are real specific LIRs who wish to have this criteria removed.

We also had voices raised in our own Policy Meeting that the "200" assignment is being a barrier to hosting providers as their customers don't match the number of assignments they make.

I hope this clarifies the background.

izumi

Philip Smith wrote:

Guangliang Pan said the following on 29/1/08 16:11:

...

Respond to a question - "How many organisations have been denied v6 space under the current proposal?"

There has been one IPv6 allocation request that did not initially include a plan for 200 assignments. An APNIC hostmaster contacted the organisation, which was able to include a plan for 200 assignments in its second email to APNIC.

Thanks Guangliang!

I suppose the question we don't know the answer to is "how many organisations simply don't bother applying for IPv6 space because they believe they MUST deploy to 200 sites within 2 years".

Note my wording, as that is the implication from the problem that the authors of the proposal want to solve.

Has anything like this been asked in any recent member survey carried out by KPMG? I'm curious, because in my extensive travels I don't find anyone who is put off because they have completely mistaken the current IPv6 allocation policy, yet both Jordi previously and the current authors are somehow implying that there is a serious problem (without supplying any evidence).

philip

•          sig-policy:  APNIC SIG on resource management policy           *
  

---

sig-policy mailing list sig-policy@lists.apnic.net http://mailman.apnic.net/mailman/listinfo/sig-policy

Sorry I haven't replied to this.

Although the current discussion is focused more on whether we need this proposal in the first place, please see my response below:

One question, why is this proposal targeted existing IPv4 holders only? I think it will be OK if 200 criteria is removed, because RIR/NIR will examine the network plan for all address requests.

I thought about that too, but it seemed to losen the criteria far more than the original intention if we simply remove the 200 criteria.

It may also increase misintepretations the other way, that any organization which makes assignments to other entity is eligible for an allocation and complain if they can't.

Having said that, I'm pretty open to modifications as long as we don't end up giving IPv6 allocations to anyone.

izumi

(Tomohiro -INSTALLER- Fujisaki) wrote:

Hi,

I basically support this proposal since as Izumi-san mentioned, hosting providers said that number is a barrier to apply the IPv6 address space in JP open policy meeting and other meetings discussing IPv6 deployment in Japan.

One question, why is this proposal targeted existing IPv4 holders only? I think it will be OK if 200 criteria is removed, because RIR/NIR will examine the network plan for all address requests.

-- Tomohiro Fujisaki

| > I suppose the question we don't know the answer to is "how many | > organisations simply don't bother applying for IPv6 space because they | > believe they MUST deploy to 200 sites within 2 years". | > | > Note my wording, as that is the implication from the problem that the | > authors of the proposal want to solve. | | Thanks for clarifying. This is indeed is the issue we are trying to | solve.(i.e,to prevent people from misunderstanding the criteria and be | reserved from make requests despite their needs.) | | We found out that out of 60 members we visited last year, more than half | of small to medium LIRs haven't bothered to apply for allocations in the | first place thinking they can't meet "200 assigments" criteria. This | includes those who wish to start deployment as a test case first and not | 100% sure to start as business, so the figure could be lower if you seek | for those solid service plan. It still did help us see that there are | real specific LIRs who wish to have this criteria removed. | | We also had voices raised in our own Policy Meeting that the "200" | assignment is being a barrier to hosting providers as their customers | don't match the number of assignments they make. | | I hope this clarifies the background. | | | izumi | | Philip Smith wrote: | > Guangliang Pan said the following on 29/1/08 16:11: | >> | >> Respond to a question - "How many organisations have been denied v6 | >> space under the current proposal?" | >> | >> There has been one IPv6 allocation request that did not initially | >> include a plan for 200 assignments. An APNIC hostmaster contacted the | >> organisation, which was able to include a plan for 200 assignments in | >> its second email to APNIC. | > | > Thanks Guangliang! | > | > I suppose the question we don't know the answer to is "how many | > organisations simply don't bother applying for IPv6 space because they | > believe they MUST deploy to 200 sites within 2 years". | > | > Note my wording, as that is the implication from the problem that the | > authors of the proposal want to solve. | > | > Has anything like this been asked in any recent member survey carried | > out by KPMG? I'm curious, because in my extensive travels I don't find | > anyone who is put off because they have completely mistaken the current | > IPv6 allocation policy, yet both Jordi previously and the current | > authors are somehow implying that there is a serious problem (without | > supplying any evidence). | > | > philip | > -- | > * sig-policy: APNIC SIG on resource management policy * | > _______________________________________________ | > sig-policy mailing list | > sig-policy@lists.apnic.net | > http://mailman.apnic.net/mailman/listinfo/sig-policy | | * sig-policy: APNIC SIG on resource management policy * | _______________________________________________ | sig-policy mailing list | sig-policy@lists.apnic.net | http://mailman.apnic.net/mailman/listinfo/sig-policy |

On 31/01/2008 10:41, "Philip Smith" pfs@cisco.com wrote:

[...]

How about simply making the IPv6 allocation minimum requirements mirror those for IPv4?

[...]

From http://www.apnic.net/policy/add-manage-policy.html, an IPv4 allocation of a /21 is made if the LIR can justify a /23 now (section 9.3), and can demonstrate a plan to use a /22 in a year.

If we translate that into IPv6 equivalents, that would be "justify a /30 now, and demonstrate a plan to use a /31 in a year to get a /32" - which is considerably more than the 200 (which is not even a /40) in the existing IPv6 policy. Do we want to go here? 200 sounds a lot easier to me! ;-)

You can mirror the concept without mirroring the details. For instance, you could introduce a system that guarantees everyone as much space as they can demonstrate they need. When an APNIC member demonstrates a need for x /48s the would be allocated a block that meets those needs. This could be done based on an HD-ratio measurement or some other 'fair scale'.

As APNIC use a sparse allocation model there would be a very good chance that a future allocation would be contiguous with the first allocation.

Regards,

Leo

On 31/1/08 8:18 PM, "Izumi Okutani" izumi@nic.ad.jp wrote:

I thought about that too, but it seemed to losen the criteria far more than the original intention if we simply remove the 200 criteria.

Isn't APNIC's role "Addressing the challenge of responsible Internet resource distribution in the Asia Pacific region"? With one of the challenges being aggregation? (This is stated in may places in APNIC's doco)

This proposal WILL end up basically giving away address space to whoever asks for it. Izumi, you must recognise this is bad, goes against everything APNIC stands for, and should not proceed.

And please, don't say "but the other RIRs...". If they jumped off a bridge... :)

Cheers, Tim.