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

رد: [AfriNIC-rpd] What is our take on the central pool IPv4exhaustion?

Hytham EL Nakhal hytham at
Sat Sep 1 22:23:23 UTC 2007

Dear All,
As usual I'm late in reply but I rely on your generosity and understanding that I was too busy last month.
Hi Bill,
Thank you for understanding our situation, and that's why we apply this proposal "Global Policy for the Allocation of the Remaining IPv4 Address Space" for all RIRs, and as you know it get consensus in LACNIC region. But we (the authors) are working to change the value of N to become 2 instead of 5 as it's the value used now by IANA for a request from any RIR. that proposal should be discussed in APNIC24 meeting next Thursday & I hope we can get consensus on it.
Bill Said> So, I encourage everyone to do whatever they can to make IPv6 a reality as soon as possible, but also not to burn any bridges behind, just yet. 
There are a lot of users in the world who won't have access to IPv6
content for quite some time, and a lot of content that won't be accessible to IPv6-only users for quite some time.  Dual-stack deployments will exist for quite a while.

I do agree with you in that thoughts
Hi Badru Ntege,


Badru wrote>(I hasten to add that this is what I'm proposing by redefine
equal distribution to have a weighting towards LACNIC and AfriNIC making
sure that we get a fare share of the V4 pie)


It's ok for N = 2 as a last allocation for each RIR


Hi Sendoro Juma

Thank you for supporting


Hi Jordi,

Jordi wrote> Also, agree, any new IPv4 resource delivered to existing or new ISP should go with a request for IPv6 space and a clear plan for the transition in a maximum of 12-18 months. 


Agree with you (and I suggested something like that in APNIC policy mailing list as a way to encourage deploying of IPv6).


>I think it is clear that people still don't understand that is STILL
POSSIBLE to do the transition with NAT and private addresses if we run out
of IPv4. IPv4 with NAT and private address space is NOT the best solution,
not the optimal one, BUT *IT WORKS*. So there is no need to try to extend
IPv4 life and will not be good at all.


ISP’s (LIRs) when requesting IPv4 from RIR; It doesn’t request it for its core network , it request it for its customers. So when a customer ask for real IPs how could ISP convince him that he could use virtual IPs and apply NAT ? of course customer request depend on his real needs for his applications.


>Developing applications with IPv6 is far more easier than with IPv4 and this provides a path for African people to do business while the rest of the world is still spending tons of dollars/euros in developing less advanced applications that traverse NATs. 

And by the way, I still haven't seen a single network where IPv6 is not supported. There is always a good walk around for every network case.


Agree with you no need for NAT as we can have real IPs, But for walk around, It’ll cost money too; as a simple example if provider has router doesn't support IPv6 so as you said he can get a Linux server and operate it as a router which will cost him the price of reliable server can stand with his 24/7 operation and will cost him a securing system to prevent hackers from attacking this Linux server ..



>In my experience, it is easier for networks in Africa and Latin America, to
move to IPv6 than in other regions where they have much bigger networks and
older equipment.

The only thing we do delaying the transition and trying to artificially
increase the availability of IPv4 in the region is damaging the development
of the region, not in the other way around.


As an addition to what Alain Patrick AINA said, then we’ll live in isolated island if we do that before big players did. for example if Google doesn’t have dual stack servers, so how could we reach it without real IPv4 ? and also all free mail servers like yahoo, hotmail, gmail,…etc.


(Well Jordi we are here in APNIC meeting and we can discuss it better face 2 face).


Hi Mark J Elkins,

Mark wrote> AfriNIC could simply make it more difficult to provide the
remainder space.
a) For totally new organizations requesting space - they need to provide IPv6 plans.
b) For organizations with existing space - the need to apply for IPv6
space and show roll-out plans with appropriate time scales before any
new allocations?


Exactly … I agree with you in doing that to encourage the deployment of IPv6 in Africa and save -as could as we can- the last allocations AfriNIC will have for new-comers. But it'll be a next phase if we could get consensus on this proposal.


Hi Frank,

Frank wrote>So  maybe 3 /8s for AfriNIC, 2 /8s for LACNIC and 1 /8 for the other 3 more
advanced regions could work better when the pool gets to 8 /8s remaining.
The policy could even provide for a return of completely unused /8 blocks
after that specified post-exhaust grace period expires.

What about 2 /8 for each RIR ?








من: rpd-bounces at بالنيابة عن Bill Woodcock
تاريخ الإرسال: السبت 8/11/2007 9:03 م
إلى: Richard at; AfriNIC Resource Policy Discussion List
الموضوع: Re: [AfriNIC-rpd] What is our take on the central pool IPv4exhaustion?

      On Sat, 11 Aug 2007, Richard Bell wrote:
    > I have not followed the debate in detail. However clearly the cost to operators of upgrading their networks are significant. Furthermore the relative growth of african networks is gaining momentum.  Since many networks rely on the second hand equipment markets to grow cost eff ectively and since africa has the smallest share of existing ipv4 allocations, why not do something radical like lobbying for afrinic to get the lions share of what's left..................

This debate is, more or less, occurring already.  Since you have to get
agreement between the five RIRs, and particularly because Latin America is
in the same boat you are, to some degree, economically and
conservation-wise, it's politically unlikely that AfriNIC would receive
"the lion's share" per se.  However, receiving an extra one or two /8s is
a possibility, if you make a strong enough case for it.

A case that I've been trying to make is similar to one you just mention:
since the cost of the minimum new IPv6-capable router is about a tenth the
annual salary of an average American, but about ten times the annual
salary of an average African, there's a huge extra cost which African
networks will encounter in moving to IPv6, and thus they need more time
for IPv6-capable equipment to "trickle down" through the resale market, et
cetera.  So far, this argument hasn't gotten much traction.

Another case that I've made with somewhat more success is that if the
African market isn't given room for future growth, it will not be able to
develop as a potentially lucrative trading-partner for businesses in other
regions.  In other words, if you don't have IP addresses, you won't be
able to do as much electronic commerce with companies in other regions in
the future.  I had not expected that argument to be more compelling than
the other one above, but you find out what works by trying them out on
people, and this one has made more converts so far.

I would advise everyone on the list to take this debate very seriously,
and think about the issues surrounding IPv4 depletion.  I'd say that there
are two categories of issue: first, how we deal with resources in the
short time between now and when the last new IPv4 addresses are depleted (for
instance, does AfriNIC get a larger share to account for future growth,
which isn't possible under current allocation policies), and second, how
we deal with the trading or reassignment of addresses _after_ depletion,
when AfriNIC has no more to give out, and they have to be moved back and
forth between users.  Will speculators be allowed to drive prices up, or
will the market be regulated as it is now, to only include qualified
recipients, and keep costs lower?

These are very active policy debates, and we have very little time to
settle on what we, collectively as a global community of Internet users,
are going to do.  So think hard, and talk this out.  Active public debate
is the only way to arrive at a good conclusion.


rpd mailing list
rpd at

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the RPD mailing list