[lacnog] IPv6 Transitional Uncertainties

Carlos A. Afonso ca en cafonso.ca
Mie Sep 14 09:12:40 BRT 2011


Caríssimo Arturo,

Estoy seguro de la excelencia del trabajo de LACNIC y de su esfuerzo
dedicado y competente en la formación de cuadros para avanzar en la
transición a IPv6, como en otras tecnologías relacionadas a la operación
de la red (DNS seguro etc). No puedo hablar por los otros RIRs pero no
tengo razones para dudar de sus calificaciones.

Me refiero exclusivamente a la cuestión de la distribución IPv4 en esta
transición, sobretodo de la gobernanza conjunta de esa transición.

[] fraterno

--c.a.

On 09/14/2011 08:58 AM, Arturo Servin wrote:
> Carlos,
> 
> On 14 Sep 2011, at 08:27, Carlos A. Afonso wrote:
> 
>> Caríssimos/as,
>>
>> Sorry to be incisive, being a sort of outsider regarding "machine-room
>> knowledge" of many details of the issue.
>>
>> Taking some large numbers: Latin America for example, with 100 million
>> households, in which all countries' governnments in one way or another
>> are promising universal broadband for 2014, both fixed and via mobile,
>> starting from a current situation in which this means triplicating, even
>> sextuplicating real IP connections. Are you sure that by 2014 we will
>> have such a full IPv6 deployment so that in this expansion we can simply
>> forget about IPv4 scarcity?
> 
> We are sure we cannot forget it. Even for the most positive scenario, IPv4 will be there.
> 
> 
>>
>> I see a situation in which every household will probably need two, three
>> real IPv4 IPs each (3G/LTE smartphones, ADSL etc) with very little
>> possibility of reuse since most will be permanent connections, unless in
>> less than three years IPv6 will be fully deployed throughtout the
>> Internet. Brazil and Argentina alone will probably need about 30-40
>> million new IPv4 addresses by 2014 -- Brazil alone has more than 50
>> million households and currently only about 12-13 million are on the
>> "broadband grid" -- not to speak of the expanding institutional IP needs
>> for business, schools etc. Or are we satisfied with the "NATting" of the
>> Internet Huston describes to force real IP sharing by the end user?
> 
> No, NAT (CGNs) is not the solution. We have said it many times, over and over. 
> 
>>
>> Maybe I am seeing ghosts where there are not, but I am far from thinking
>> we should not worry.
> 
> You are not seeing ghosts, you should be worried. We all are.
> 
>> And, yes, I continue to think RIRs are too lame
>> about it.
> 
> RIRs we do not provide IPv6 access. The problem are not RIRs, not even content (Google, Yahoo, MS, Facebook are ready), it is access and CPEs. 
> 
> And for the record, we RIRs (at least LACNIC) provide (and have provided) IPv6 training (virtual seminars, workshops), we give (have given) many talks about this (IPv6 and IPv4 exhaustion and governance), we have our infrastructure in v6 (www, whois, etc.), we have our office full dual-stack, we do not buy Internet link if the ISP does not provide native v6, etc., etc. 
> 
> So, we could be anything, but lame about IPv6 and IPv4 exhaustion is not one.
> 
>> Please do convince me I am absolutely wrong!
> 
> Our actions speak for ourselves.
> 
> Regards,
> .as
> 
>>
>> --c.a.
>>
>> On 09/14/2011 12:34 AM, Fernando Gont wrote:
>>> Raul,
>>>
>>> Please find my comments in-line...
>>>
>>> On 09/13/2011 11:07 PM, Raul Echeberria wrote:
>>>>
>>>> It is not true that the other RIRs, as you said, are happy with the
>>>> inequity in the allocations of IPv4 addresses. LACNIC has stated very
>>>> clearly many times our position regarding legacy addresses, those
>>>> addresses that were allocated before the existence of the RIRs. In
>>>> fact I spoke about that in the last LACNIC meeting in Cancun in my
>>>> report.
>>>
>>> I was just about to mention your presentation at the past LACNIC meeting
>>> in Cancun, but wasn't able to find the video of it in Youtube to provide
>>> a pointer. Is it online somewhere?
>>>
>>>
>>>> What is very important from Geoff's article is that we have to learn
>>>> the lessons from APNIC region's experience. in our region we have a
>>>> good stock of IPv4 addresses to support a smooth transition to IPv6,
>>>> but if the Network operators don't take the necessary measures very
>>>> soon, so in a couple of years we will be in the same situation than
>>>> APNIC region is now.
>>>
>>> FWIW, I'm personally planning to submit a policy to the "politicas"
>>> mailing-list proposing that new IPv4 blocks are assigned
>>> inversely-proportional to the number of IPv4 addresses that they
>>> currently "own".
>>>
>>> That means that those who are making more money (and are also likely to
>>> be in the business for longer), are the ones that will have to stop
>>> consuming IPv4 address from the pool, and focus on deploying v6.
>>>
>>> That also means that there will be IPv4 addresses for new players.
>>>
>>> Thanks,
>> _______________________________________________
>> LACNOG mailing list
>> LACNOG en lacnic.net
>> https://mail.lacnic.net/mailman/listinfo/lacnog
> 
> _______________________________________________
> LACNOG mailing list
> LACNOG en lacnic.net
> https://mail.lacnic.net/mailman/listinfo/lacnog
> 



Más información sobre la lista de distribución LACNOG