Search RPD Archives
Limit search to: Subject & Body Subject Author
Sort by:

[rpd] Re: [afnog] IPv4 Allocations by Length Statistics for 2014

Adam Nelson adam at
Fri Nov 7 07:29:10 UTC 2014


You're right and I 100% agree with your sentiment.  Being totally open
isn't always realistic and can be counterproductive.

With that said, operators should be educated that their competitors know
alot about them from closed-source sources (Dyn Corp as mentioned
previously).  If their big fear is public disclosure, they should first
confirm what's available privately.  Sometimes, making data public
undermines the private sources and can be good corporate policy (i.e. good
for valuation, good for self-discovery in large organizations that are
balkanized, good for recruiting).

Also, we should keep in mind that other RIRs are doing research - they're
just not using information that's under NDA.  RIPE has a ton of probes in
Africa sucking up data.  Here's mine for instance:!tab-network


Kili - Cloud for Africa:
Musings: <>
More Musings:
About Adam:

On Thu, Nov 6, 2014 at 8:54 PM, Owen DeLong <owen at> wrote:

> It feels weird to be on the side of this debate that I am about to
> represent, however…
> I think this discussion has strayed far from the mission of this mailing
> list. I think we should ask ourselves the following questions before
> continuing this thread along its current lines:
> 1. What data should the RIR be able to collect in order to do its job?
> 2. Is collecting that data more likely to encourage policy compliance or
> more likely to encourage fraud?
> 3. Of the data we (as an RIR) collect, how much of it should the RIR be
> required to disclose to the public?
> Greater transparency is generally good on the surface, but the more of the
> data you ask for that people don’t want disclosed,
> the greater your risks in question 2. This goes up even more if you don’t
> build confidence that the data will be kept confidential.
> I think we should choose which data is collected strictly in terms of data
> that enables the RIR to issue resources in compliance with policy with a
> reasonable ability to detect fraud. We should only disclose that
> information which is vital to other operators and users of the internet for
> contact purposes to address problems (abuse, failures, etc.). We should not
> require disclosure of confidential information that must be collected in
> order for the RIR to be able to reasonably audit or analyze requests for
> policy compliance as this will reduce the probability of getting accurate
> data and/or voluntary compliance with policies.
> I generally favor sunshine as the best disinfectant and greater
> transparency, especially in organizations of public trust. However, it
> seems to me that in this case, forced disclosure is more likely to increase
> fraud than prevent it.
> If we want to fund wider research on a voluntary basis, I’m all for that,
> but in terms of required information for the RIR to process requests and
> mandatory disclosures, I think we must be very careful not to do harm to
> the core mission.
> Further, unless it is the basis for number resource policy development, I
> think this list may not be the correct forum for developing said research.
> These are just my opinions and you are certainly welcome to tell me that I
> am wrong. I have no standing to tell anyone what they should be doing and
> that is not my intent here. I merely ask that you consider these issues in
> the context of allowing the RIR to fulfill its core mission of managing
> address space.
> Owen
> On Nov 6, 2014, at 4:04 AM, Adam Nelson <adam at> wrote:
> I've never been to Nigeria but SA is a highly developed country with
> nuclear power plants, dozens of world class airports, dual carriageways,
> etc...
> I would say, presuming those numbers are even correct, that it's simply
> Internet maturity reflecting human development generally.  Maybe the better
> comparison would be Kenya and Nigeria or Angola and Nigeria.
> --
> Kili - Cloud for Africa:
> Musings: <>
> More Musings:
> About Adam:
> On Thu, Nov 6, 2014 at 2:27 PM, Walubengo J <jwalu at> wrote:
>> @Adam,
>> And perhaps Dyncorp (or similar research institution) would be able to
>> tell us why Nigeria (largest economy in Africa) is enjoying, without
>> complaining only 3% total share of the IPv4 resources in Africa while SA,
>> the second largest economy is doing 38%.
>> Perhaps they are doing too much NATting or worse still "re-using" IP nos?
>> Maybe Sunday would know the reasons :-).
>> walu.
>> --------------------------------------------
>> On Thu, 11/6/14, Adam Nelson <adam at> wrote:
>>  Subject: [rpd] Re: [afnog] IPv4 Allocations by Length Statistics for 2014
>>  To: "Mark Tinka" <mark.tinka at>
>>  Cc: "rpd List" <rpd at>, "afnog" <afnog at>
>>  Date: Thursday, November 6, 2014, 11:39 AM
>>  Even if
>>  operators were inclined to give the data, I doubt you'd
>>  get much participation since it's all hassle and no
>>  gain.
>>  The only way to get
>>  the data is good old fashioned research - which Afrinic is
>>  probably not in a position to do internally (for various
>>  reasons including the NDA).  I bet Dyncorp (the link I sent
>>  before) has quite a bit of the information you're
>>  looking for - but it won't be cheap.
>>  -Adam
>>  --Kili - Cloud for
>>  Africa:
>>  Musings: Musings: varud.comAbout Adam:
>>  On Thu, Nov 6, 2014 at
>>  11:34 AM, Mark Tinka <mark.tinka at>
>>  wrote:
>>  On
>>  Thursday, November 06, 2014 10:07:09 AM Kofi ANSA AKUFO
>>  wrote:
>>  > Could we get more information from the data as to
>>  what
>>  > technology and applications are consuming the
>>  resources?
>>  I'm not sure operators would be willing to openly
>>  share that
>>  data, Kofi.
>>  In fact, any such discussions between operators and
>>  are under NDA.
>>  Mark.
>>  -----Inline Attachment Follows-----
>>  _______________________________________________
>>  rpd mailing list
>>  rpd at
> _______________________________________________
> rpd mailing list
> rpd at
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the RPD mailing list