Send ARIN-PPML mailing list submissions to
[email protected]
To subscribe or unsubscribe via the World Wide Web, visit
http://lists.arin.net/mailman/listinfo/arin-ppml
or, via email, send a message with subject or body 'help' to
[email protected]
You can reach the person managing the list at
[email protected]
When replying, please edit your Subject line so it is more specific
than "Re: Contents of ARIN-PPML digest..."
Today's Topics:
1. Re: IPv6 as justification for IPv4? (Owen DeLong)
2. Re: IPv6 as justification for IPv4? (David Farmer)
3. Re: IPv6 as justification for IPv4? (John Springer)
4. Re: IPv6 as justification for IPv4? (John Springer)
5. Re: Justifying an ISP /22 (Otis L. Surratt, Jr.)
6. Re: Justifying an ISP /22 (Randy Carpenter)
7. Re: Justifying an ISP /22 (Otis L. Surratt, Jr.)
----------------------------------------------------------------------
Message: 1
Date: Wed, 17 Apr 2013 09:59:46 -0600
From: Owen DeLong <[email protected]>
To: William Herrin <[email protected]>
Cc: John Curran <[email protected]>, ARIN-PPML List
<[email protected]>
Subject: Re: [arin-ppml] IPv6 as justification for IPv4?
Message-ID: <[email protected]>
Content-Type: text/plain; charset=us-ascii
On Apr 16, 2013, at 10:08 PM, William Herrin <[email protected]> wrote:
> On Tue, Apr 16, 2013 at 10:09 PM, John Curran <[email protected]> wrote:
>> The ISP must demonstrate the need for IPv4 address resources under
>> ARIN standard allocation policies (i.e. a single-homed ISP showing
>> need for a /20 or multi-homed ISP showing need for a /22) in order
>> to qualify to receive resources via transfer. Once qualified, we
>> can approve the transfer of IPv4 space; this can be to a maximum
>> of their documented need based on their current utilization rate
>> extended 24 months out, and down to a minimum of a single /24 (as
>> /24 is the explicit minimum transfer size specified in NRPM 8.3)
>
> Thanks John.
>
> So, what would folks think of a policy adjustment along these lines:
>
> "Add to: 8.3 Conditions on recipient of the transfer:
>
> * Minimum address block size qualifications defined in section 4 do
> not apply to transfers to specified recipients."
What problem do you think that would solve?
In the current case being discussed, it isn't the block size minimums he
is having a problem with, it is the amount he can get SWIPd vs. the
inability to qualify under immediate need.
IMHO, the correct fix is to modify the pre-existing space requirements
so as to allow documented need to substitute for SWIP'd space.
Owen
------------------------------
Message: 2
Date: Wed, 17 Apr 2013 11:38:01 -0500
From: David Farmer <[email protected]>
To: William Herrin <[email protected]>
Cc: ARIN-PPML List <[email protected]>
Subject: Re: [arin-ppml] IPv6 as justification for IPv4?
Message-ID: <[email protected]>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
On 4/17/13 10:59 , Owen DeLong wrote:
>
> On Apr 16, 2013, at 10:08 PM, William Herrin <[email protected]> wrote:
>
>> On Tue, Apr 16, 2013 at 10:09 PM, John Curran <[email protected]> wrote:
>>> The ISP must demonstrate the need for IPv4 address resources under
>>> ARIN standard allocation policies (i.e. a single-homed ISP showing
>>> need for a /20 or multi-homed ISP showing need for a /22) in order
>>> to qualify to receive resources via transfer. Once qualified, we
>>> can approve the transfer of IPv4 space; this can be to a maximum
>>> of their documented need based on their current utilization rate
>>> extended 24 months out, and down to a minimum of a single /24 (as
>>> /24 is the explicit minimum transfer size specified in NRPM 8.3)
>>
>> Thanks John.
>>
>> So, what would folks think of a policy adjustment along these lines:
>>
>> "Add to: 8.3 Conditions on recipient of the transfer:
>>
>> * Minimum address block size qualifications defined in section 4 do
>> not apply to transfers to specified recipients."
>
> What problem do you think that would solve?
>
> In the current case being discussed, it isn't the block size minimums he
> is having a problem with, it is the amount he can get SWIPd vs. the
> inability to qualify under immediate need.
>
> IMHO, the correct fix is to modify the pre-existing space requirements
> so as to allow documented need to substitute for SWIP'd space.
>
> Owen
I'll just add; The new PDP requires Policy Proposals to have "a clear
statement of the problem with existing Internet number resource policy".
I'd like to suggest, with the new PDP it may be more fruitful for us
to change our habits and try to focus pre-proposal discussions on honing
a common understanding of the problem statement rather than honing
policy text as we have frequently done in the past.
I believe that if we clearly define the problem in its correct context
then finding the correct text to solve that problem will likely be much
easier. I'm not saying we shouldn't discuss policy text, but that
policy text without a clear problem statement may not be helpful. That
said, good policy text can sometimes help people understand the problem.
So, I believe there is a problem here, I'm just not sure we have a clear
definition of the problem that is trying to be solved, yet. And, in
this case the policy text you propose isn't helping me understand the
problem any better, if anything it has confused the problem for me.
Thanks
--
================================================
David Farmer Email: [email protected]
Office of Information Technology
University of Minnesota
2218 University Ave SE Phone: 1-612-626-0815
Minneapolis, MN 55414-3029 Cell: 1-612-812-9952
================================================
------------------------------
Message: 3
Date: Wed, 17 Apr 2013 10:48:15 -0700 (PDT)
From: John Springer <[email protected]>
To: Owen DeLong <[email protected]>
Cc: ARIN-PPML List <[email protected]>
Subject: Re: [arin-ppml] IPv6 as justification for IPv4?
Message-ID: <[email protected]>
Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
On Wed, 17 Apr 2013, Owen DeLong wrote:
>
> In the current case being discussed, it isn't the block size minimums he
> is having a problem with, it is the amount he can get SWIPd vs. the
> inability to qualify under immediate need.
>
> IMHO, the correct fix is to modify the pre-existing space requirements
> so as to allow documented need to substitute for SWIP'd space.
>
> Owen
>
Agree
John Springer
------------------------------
Message: 4
Date: Wed, 17 Apr 2013 10:55:24 -0700 (PDT)
From: John Springer <[email protected]>
To: David Farmer <[email protected]>
Cc: ARIN-PPML List <[email protected]>
Subject: Re: [arin-ppml] IPv6 as justification for IPv4?
Message-ID: <[email protected]>
Content-Type: TEXT/PLAIN; format=flowed; charset=US-ASCII
Inline
On Wed, 17 Apr 2013, David Farmer wrote:
> On 4/17/13 10:59 , Owen DeLong wrote:
>>
>> On Apr 16, 2013, at 10:08 PM, William Herrin <[email protected]> wrote:
>>
>>> On Tue, Apr 16, 2013 at 10:09 PM, John Curran <[email protected]> wrote:
>>>> The ISP must demonstrate the need for IPv4 address resources under
>>>> ARIN standard allocation policies (i.e. a single-homed ISP showing
>>>> need for a /20 or multi-homed ISP showing need for a /22) in order
>>>> to qualify to receive resources via transfer. Once qualified, we
>>>> can approve the transfer of IPv4 space; this can be to a maximum
>>>> of their documented need based on their current utilization rate
>>>> extended 24 months out, and down to a minimum of a single /24 (as
>>>> /24 is the explicit minimum transfer size specified in NRPM 8.3)
>>>
>>> Thanks John.
>>>
>>> So, what would folks think of a policy adjustment along these lines:
>>>
>>> "Add to: 8.3 Conditions on recipient of the transfer:
>>>
>>> * Minimum address block size qualifications defined in section 4 do
>>> not apply to transfers to specified recipients."
>>
>> What problem do you think that would solve?
>>
>> In the current case being discussed, it isn't the block size minimums he
>> is having a problem with, it is the amount he can get SWIPd vs. the
>> inability to qualify under immediate need.
>>
>> IMHO, the correct fix is to modify the pre-existing space requirements
>> so as to allow documented need to substitute for SWIP'd space.
>>
>> Owen
>
> I'll just add; The new PDP requires Policy Proposals to have "a clear
> statement of the problem with existing Internet number resource policy". I'd
> like to suggest, with the new PDP it may be more fruitful for us to change
> our habits and try to focus pre-proposal discussions on honing a common
> understanding of the problem statement rather than honing policy text as we
> have frequently done in the past.
>
> I believe that if we clearly define the problem in its correct context then
> finding the correct text to solve that problem will likely be much easier.
> I'm not saying we shouldn't discuss policy text, but that policy text without
> a clear problem statement may not be helpful. That said, good policy text
> can sometimes help people understand the problem.
>
> So, I believe there is a problem here, I'm just not sure we have a clear
> definition of the problem that is trying to be solved, yet. And, in this
> case the policy text you propose isn't helping me understand the problem any
> better, if anything it has confused the problem for me.
Paste from arin-discuss, from an exchange between Owen and Randy Carpenter
>> I think the real problem here is requiring pre-existing PA space of
>> certain amounts as the initial test. The combination of a customer
>> base, need, and efficient utilization of any PA space is probably the
>> better test.
> This is something that I believe really needs fixed (and needs to be
> fixed very quickly).
> -Randy
This seems to be rather far along to a problem statement. I agree with
both statements.
John Springer
------------------------------
Message: 5
Date: Wed, 17 Apr 2013 16:41:03 -0500
From: "Otis L. Surratt, Jr." <[email protected]>
To: <[email protected]>
Subject: Re: [arin-ppml] Justifying an ISP /22
Message-ID: <[email protected]>
Content-Type: text/plain; charset="us-ascii"
Jumping on ppml late..
Randy,
>1. Be an ISP with *any* amount of space from an upstream
I'm not so sure this would be advisable? Wouldn't it be better to have
at least a /24? Or is this what you had in mind?
>2. Be able to justify a 3-month need for a /22
>3. Have upstream ISPs that refuse to hand out any more PA space (which
appears to be the rule rather than the exception now) (I'm not sure how
one would go about proving this, but I have seen situations where the
upstream would likely write a letter of support)
>4. Be forced to take an automatic IPv6 allocation (at whatever
NRPM-supported size is appropriate (preferably /32 min.))
I agree with the /32 minimum.
I remember when we became an ARIN member we had 1 /24 PA and 1 /23 PA
plus we were multi-homed so we could justify the /22.
We actually had an ASN first then we got our first allocation about 1
month later. Either way, renumbering was not that bad but still could
have been avoided with what you are proposing.
- Otis
------------------------------
Message: 6
Date: Wed, 17 Apr 2013 18:11:45 -0400 (EDT)
From: Randy Carpenter <[email protected]>
To: "Otis L. Surratt, Jr." <[email protected]>
Cc: [email protected]
Subject: Re: [arin-ppml] Justifying an ISP /22
Message-ID:
<[email protected]>
Content-Type: text/plain; charset=utf-8
----- Original Message -----
> Jumping on ppml late..
>
> Randy,
>
> >1. Be an ISP with *any* amount of space from an upstream
>
> I'm not so sure this would be advisable? Wouldn't it be better to have
> at least a /24? Or is this what you had in mind?
A /24 would not be too much of a hurdle, so that would be fine. Most of the
frustrating examples I have seen are small ISPs that have a /22 or a /21, but
are not multi-homed. They are expanding to the point of needing more space,
e.g. they have a /21, but need a /20. They don't already have a /20, so they
can't get the /20. The upstream ISP will not issue them any more space. I have
seen this on several occasions.
Some have chosen to get multi-homed, but for some, multi-homing (at least to
different upstreams) is not possible.
Looking at the NRPM more closely, I think it would be pretty a simple to have a
policy change that simply reduced the amount of PA space required for a
single-homed ISP to /24, but also require justification for a full /22.
Basically, even the playing field for single-home and multi-home, since there
is now a situation where ARIN has available address space, but upstream ISPs
seem not to have (or unwilling to provide).
The one situation that is still left hanging by this is a brand new ISP, which
can not get any PI space at all, since they have no upstream space yet. It may
be as hard for them to get new PA space as it is existing ISPs to get
additional PA space.
I understand the intent of the current NRPM, which is to prevent every Joe
Wannabe with a day-old certificate of incorporation from snagging up space that
they may never actually use. However, when if reach the point that upstream
ISPs across the board refuse to hand out any space at all, and there is still
space in ARIN's free pool, I think that may need to be revisited as well. I
don't think we are quite there yet (assuming we make the previously mentioned
changes to only require a /24 of PA space).
-Randy
------------------------------
Message: 7
Date: Wed, 17 Apr 2013 17:49:08 -0500
From: "Otis L. Surratt, Jr." <[email protected]>
To: "Randy Carpenter" <[email protected]>
Cc: [email protected]
Subject: Re: [arin-ppml] Justifying an ISP /22
Message-ID: <[email protected]>
Content-Type: text/plain; charset="utf-8"
Reply inline
-----Original Message-----
From: Randy Carpenter [mailto:[email protected]]
Sent: Wednesday, April 17, 2013 5:12 PM
To: Otis L. Surratt, Jr.
Cc: [email protected]
Subject: Re: [arin-ppml] Justifying an ISP /22
----- Original Message -----
> Jumping on ppml late..
>
> Randy,
>
> >1. Be an ISP with *any* amount of space from an upstream
>
> I'm not so sure this would be advisable? Wouldn't it be better to have
> at least a /24? Or is this what you had in mind?
>A /24 would not be too much of a hurdle, so that would be fine. Most of the
>frustrating examples I have seen are small ISPs that have a /22 or a /21, but
>are not multi-homed. They are expanding to the point >of needing more space,
>e.g. they have a /21, but need a /20. They don't already have a /20, so they
>can't get the /20. The upstream ISP will not issue them any more space. I have
>seen this on several >occasions.
I've seen this as well. In particular, was involving moving from /20 range to
/19.
That seems to be the case with many of the stories I hear.
>Some have chosen to get multi-homed, but for some, multi-homing (at least to
>different upstreams) is not possible.
This is always a tough one. Because it costs money to bring in a separate pipe.
But one could simple fix this by colocating gear in a neutral carrier data
center and speaking BGP. By use of PTP or tunnels to get it back to your POP.
But that defeats the point of redundancy but it achieves multi-homing to
satisfy policy. But couldn't those ISPs with those issues simply explain their
situation and I'm sure ARIN could work with them?
>Looking at the NRPM more closely, I think it would be pretty a simple to have
>a policy change that simply reduced the amount of PA space required for a
>single-homed ISP to /24, but also require justification >for a full /22.
>Basically, even the playing field for single-home and multi-home, since there
>is now a situation where ARIN has available address space, but upstream ISPs
>seem not to have (or unwilling to >provide).
But couldn't you just adjust the language to account for dual-homing to the
single ISP?
The /24 PA would help out.
>The one situation that is still left hanging by this is a brand new ISP, which
>can not get any PI space at all, since they have no upstream space yet. It may
>be as hard for them to get new PA space as it is >existing ISPs to get
>additional PA space.
This can be a tough one... But it could be fixed by having the applicant submit
copies of contract with an ISP(s) and their request for PA. This is the same
approach that is done for ASNs I believe if you are not yet multihomed.
------------------------------
_______________________________________________
ARIN-PPML mailing list
[email protected]
http://lists.arin.net/mailman/listinfo/arin-ppml
End of ARIN-PPML Digest, Vol 94, Issue 37
*****************************************