[arin-ppml] ARIN-prop-176 Increase Needs-Based Justification to 60 months on 8.3 Specified Transfers

Owen DeLong owen at delong.com
Sat Jun 30 05:23:50 EDT 2012


On Jun 29, 2012, at 5:56 PM, Jimmy Hess wrote:

> On 6/29/12, Milton L Mueller <mueller at syr.edu> wrote:
>> oops. You may not realize it, but you just admitted that there is no such
>> thing as "the community." You've observed that the people involved here fall
>> into a variety of different groupings with different interests - legacy
>> holders, speculators, prospective buyers who "want more resources than they
>> can justify," and others, good guys like you and Hurricane and Comcast.
> 
> Let's be really really clear.  The ARIN community consists of
> Internet Service Providers and network operators,  that require
> number resources for the purpose of connecting their networks
> together,  using IP, and require the WHOIS service to help coordinate
> with other networks, for reliable connectivity, and minimal network
> abuse.    If you're anyone else, you are not a member of the community
> that it's ARIN's job to provide services to.
> 

This simply isn't true. The community is much more encompassing than
those that ARIN provides registration services to.

The community is literally anyone with an interest in IP address policy.

The voting membership (those that elect the AC and the Board) are ARIN
members that hold resources.

One need not be a member of ARIN in order to be a member of the community.
One need not hold resources to be a member of the community.
One need not operate a network to be a member of the community.

If you are signed up on PPML and participate, you are a member of the
community.

> If you happen to by coincidence  be both a speculator AND an ISP, or
> both a prospective buyer, and the operator of a network.
> 
> Then you are a member of the community,  but your special interests or
> roles as "speculator",   "prospective buyer",  "academic", or
> whatever,  have nothing to do with the community, and should be left
> at home.

For better or worse, speculators, hoarders, derivatives traders, etc. are
actually members of the community if they choose to participate.

> Although you're certainly allowed to share your opinions and views,
> you should ensure that you disclose any special financial or personal
> interest you have.

That would be nice, but to the best of my knowledge, it is only required
of AC and Board members where there is a clear conflict of interest policy
that would apply.

> And ARIN in providing proper stewardship of the resources,  should be ensuring
> that factions except "network operator"' organizations (and network end user)
> organizations'   are not being served,

Again, not true. ARIN should be ensuring that those factions do not receive
resource registration services (unless the community decides to permit them
to receive resource registration services through policy changes). However,
they are entitled to the services of participation in the policy development
process just as any network operator, ISP, end user, etc.

> Especially not at the expense of ISPs and network end users  best interests
> specifically in those roles.

Who gets to decide what that best interest is and who qualifies and who doesn't?
What are the criteria for such a decision?


> And that should be true,  even if  non-Network operator  factions
> become more numerous
> on PPML and policy meetings,   due to the fact the financial
> incentive,  might encourage
> more aggressive participation,  in order to manipulate policy.

As much as I fear this eventuality and it is the primary reason that I am not
in favor of pushing 8.3 further down the slippery slope, I'm not sure how we
can remain true to our charter and bylaws and the process we have created
and actually prevent this.

> ARIN should be sure to recognize and reject any manipulation efforts of that
> nature that are not in the interest of the specific  community ARIN
> exists to serve.

ARIN who? ARIN staff? The ARIN community (in which case, by what criteria
would you define the portion of the community that gets to make such decisions)?
It's all well and good to say "ARIN", but, ARIN isn't a body that takes action per se.

>> As my previous message indicated, I have a very strong suspicion that the 24
>> month period is arbitrary and has no more or less support than 60 months or
>> 12 months or 6 months. But since you are saying that the data does not
> 
> The 24 month period is surely an arbitrary number that was chosen, but there
> was some agreement to, but the contention that it  is no more/less
> supported than 60 months  is wrong.

True. There was quite a bit of opposition to 36 or more months when that
question was asked at the last PPM in Vancouver whereas there was quite
a bit less opposition to 24 months (though I felt there was significant
opposition there as well).

>> Tell me: what _kind_ of data would indicate to you that a change is
>> warranted? Would it be something like a yield curve for address blocks?
> 
> There are three things that all need to be shown simultaneously,
> before there's any reason to believe the justification period is too
> short:
> 
> 1. Proof of an unmet need for IP addresses in the region, for the
> purpose of actually running networks, per the RFC2050 criteria;  there
> must be organizations actively seeking resources who are not getting
> them,  because there are not enough resources offered by specified
> transfer.
> 
> 2. Lack of interest by organizations seeking IPv4 resources through
> specified transfer
> OR  organizations that have unused resources leaving them idle and
> neither returning
> to ARIN nor offering by specified transfer  (an actual showing that
> organizations really are not participating in the transfer market due
> to the inefficiency caused by the requirement  that they execute
> multiple transfers instead of just one).

I'm not sure how the first criteria proves that a longer justification period
is needed. Afterall, at a certain point in the (hopefully) not too distant
future, it might merely indicate that IPv4 is no longer expanding.

> 3.   That there is no  equally expensive, less expensive, or more
> familiar/convenient source of IP addresses than specified transfer.
> Until (3) is met, there is no possible data that will show that a
> longer justification period is suitable, justified, necessary, or
> appropriate.

Agreed.

> For example:  as long as there are free pool resources  available at
> similar cost,  Transfer is a less attractive option,  because  direct
> allocation from a pool is adequate,  and with a high likelihood
> provides  "Virgin"   IP addresses   that have never been allocated
> before, and probably aren't on blacklists.

This is not necessarily true of all free-pool allocations/assignments.
Some resources in the freepool are, indeed, recycled.

> Also, as long as there is still a free pool,  organizations in the
> ARIN region  that have dealt with ARIN in the past are likely to
> continue to make their additional resource requests  through ARIN,
> because they have dealt with ARIN in the past,  there is a degree of
> familiarity and trust.

Yes, though the 3/24 month dichotomy with transfers does put a bit
of a damper on that.

> Transfer resources must be significantly less expensive, before many
> organizations will consider taking on added risk of dealing with
> potentially unfamiliar third parties,  and lengthening the process of
> getting their resources. They may even prefer to acquire other orgs
> altogether,  and perform 8.2 transfers instead.

Another thing worth considering is that the transfer market is currently
somewhere between $10 and $12 per address IIRC. Only organizations
with sufficient revenue to support an ROI on that price would consider
purchasing them at that price. Many ISPs and other network operators
are low-margin businesses and it is entirely possible that expanding
their IPv4 customer base would not provide a sufficient ROI to continue
expanding with that cost model.

> As long as there is a free pool, the data is expected to be skewed,
> and therefore
> unreliable.

Yep.

Owen




More information about the ARIN-PPML mailing list