[arin-ppml] quantitative study of IPv4 address market

[Mike Burns]

I think Milton's analysis was ingenious and the fact that more than a year
out Microsoft is routing less than 9,000 out of the 660,000 addresses
purchased from Nortel all but puts the lie to the idea that an accurate
justification was made.
The idea that Microsoft paid $7.5 million because their need was so acute
that processing successive 3-month justifications from the free pool would
be too risky is ludicrous when viewing the paucity of routed addresses more
than 12 months out.
There are over 17 million RFC 1918 addresses that can be privately used!

Can anybody speculate on a valid justification for the delivery of 660,000
addresses required to be used within a year, with 98% of them to be
unrouted?

It's laughable.

My interpretation is that ARIN staff knew that with the privacy rights
attending to the LRSA and the needs test, they could bluster their way
through.
I don't think they considered the kind of post-hoc analysis Milton
performed, and are gnashing their teeth at Microsoft for so obviously
showing them up.

Kudos to Prof. Mueller for pulling back the curtain a bit.

Regards,

Mike Burns






-----Original Message----- 
From: John Curran
Sent: Wednesday, September 05, 2012 12:08 PM
To: Milton L Mueller
Cc: arin-ppml at arin.net
Subject: Re: [arin-ppml] quantitative study of IPv4 address market

On Sep 5, 2012, at 4:39 PM, Milton L Mueller <mueller at syr.edu> wrote:

> [Milton L Mueller] John, you probably read an earlier version of the study
> which did not include our analysis of the use of the number blocks
> purchased by MSFT. (We wanted to verify certain methodological aspects of
> that before publishing it.)
>
> We found that since acquiring the addresses, which is now more than a year
> ago, MSFT has actually _reduced_ the number of blocks being routed. They
> have only put into service a few /24s and /22s I think, and have actually
> withdrawn from use a couple of /16s.

Milton -

   Use of IP address blocks is independent of routing of address blocks,
   particularly if you operate data centers with numerous servers.  The
   desire for using globally unique numbering for such devices often does
   not result in the address blocks being globally routed.

> Now, I cannot assert that ARIN did not do a needs assessment. Only ARIN
> staff can verify or refute that.


   It was done, and you have been told that by ARIN staff (i.e. myself) on
   numerous occasions, and yet you continue to publish allusions to the
   contrary.  I understand this as a journalistic technique, but not as any
   form of academic research.

> I can, however, assert with confidence that a 1-year time horizon was used
> at that time (according to your policy in place), and that the entity who
> claimed need based on that timeline is not, after more than one year,
> using the vast majority of the addresses; indeed, fewer addresses from the
> Nortel block are being routed after the transaction than before.
> Therefore, there are legitimate questions to be raised about what “needs
> assessment” actually consists of.

   Alas, assuming that address block usage only occurs with the public
routing
   of the address block is a flawed methodology.

FYI,
/John

John Curran
President and CEO
ARIN




_______________________________________________
PPML
You are receiving this message because you are subscribed to
the ARIN Public Policy Mailing List (ARIN-PPML at arin.net).
Unsubscribe or manage your mailing list subscription at:
http://lists.arin.net/mailman/listinfo/arin-ppml
Please contact info at arin.net if you experience any issues.

-----Original Message----- 
From: John Curran
Sent: Wednesday, September 05, 2012 12:08 PM
To: Milton L Mueller
Cc: arin-ppml at arin.net
Subject: Re: [arin-ppml] quantitative study of IPv4 address market

On Sep 5, 2012, at 4:39 PM, Milton L Mueller <mueller at syr.edu> wrote:

> [Milton L Mueller] John, you probably read an earlier version of the study 
> which did not include our analysis of the use of the number blocks 
> purchased by MSFT. (We wanted to verify certain methodological aspects of 
> that before publishing it.)
>
> We found that since acquiring the addresses, which is now more than a year 
> ago, MSFT has actually _reduced_ the number of blocks being routed. They 
> have only put into service a few /24s and /22s I think, and have actually 
> withdrawn from use a couple of /16s.

Milton -

   Use of IP address blocks is independent of routing of address blocks,
   particularly if you operate data centers with numerous servers.  The
   desire for using globally unique numbering for such devices often does
   not result in the address blocks being globally routed.

> Now, I cannot assert that ARIN did not do a needs assessment. Only ARIN 
> staff can verify or refute that.


   It was done, and you have been told that by ARIN staff (i.e. myself) on
   numerous occasions, and yet you continue to publish allusions to the
   contrary.  I understand this as a journalistic technique, but not as any
   form of academic research.

> I can, however, assert with confidence that a 1-year time horizon was used 
> at that time (according to your policy in place), and that the entity who 
> claimed need based on that timeline is not, after more than one year, 
> using the vast majority of the addresses; indeed, fewer addresses from the 
> Nortel block are being routed after the transaction than before. 
> Therefore, there are legitimate questions to be raised about what “needs 
> assessment” actually consists of.

   Alas, assuming that address block usage only occurs with the public 
routing
   of the address block is a flawed methodology.

FYI,
/John

John Curran
President and CEO
ARIN




_______________________________________________
PPML
You are receiving this message because you are subscribed to
the ARIN Public Policy Mailing List (ARIN-PPML at arin.net).
Unsubscribe or manage your mailing list subscription at:
http://lists.arin.net/mailman/listinfo/arin-ppml
Please contact info at arin.net if you experience any issues.