[rpd] Questions for Alain...

[Owen DeLong]

> On Jun 7, 2018, at 12:15 , Noah <noah at neo.co.tz> wrote:
> 
> 
> 
> On Thu, 7 Jun 2018, 5:30 p.m. Owen DeLong, <owen at delong.com <mailto:owen at delong.com>> wrote:
> 
> 
>> So in otherwords we should never give audience to any Inbound or Outbound or Inter-RIR IPv4 transfer policies ever.
> 
> Not what I’m saying at all. IMHO, a bidirectional inter-RIR transfer policy which takes efFect upon free pool runout would not be an unreasonable step.
> 
> Owen
> 
> Now you are changing tunes. Which RIR has reasonable space to make available to AfriNIC who with LACNIC never got the lion share in the first place.

Nope… I haven’t changed my tune about this at all.

Transfers are not about space held by RIRs. They are about space held by organizations that are willing to part with it.

> Inter-RIR is not an option. We dont have space for reciprocal exchange and IPv6 is our way out and the so called Inter-RIR will slow greenfielding of IPv6.

We can agree to disagree about this. I don’t believe that inter-RIR transfers will change the rate of IPv6 greenfielding.

There have been more IPv4 inter-RIR transfers in the RIPE region than anywhere else, IIRC, yet theirs is the region with the fastest IPv6 adoption progress at this time.

The facts on the ground do not support your assertion.

> Others have different opinions. 
> 
> 
> No Inter-RIR, no Inbound and no Outbound and we move forward.
> 
> We have a policy in place to aid inside Africa transfer of space and this will work for us in the meantime.

Meh. Maybe it will, maybe it comes up short. Time will certainly tell.

To me, transfer policies are ways to enable people who want to do so every opportunity to rearrange the deck chairs.

The IPv4 titanic is still sinking, regardless of where the chairs end up.

>> We should also never allow any more IPv4 space into this continent.
> 
> There is a huge difference between preserving the appearance of a free pool through policy manipulation
> 
> 
> The free pool was meant to be preserved into order to aid IPv6 transitions. So if some members believe that fast tracking IPv4 into depletion is best option to transitioning to IPv6, then be it.

The idea that preserving the free pool somehow aids IPv6 transition is laughable in the extreme.

As long as people perceive that there is an ability to go to an RIR and get more IPv4 space, they will procrastinate the deployment of IPv6 as an alternative. This behavior is clearly demonstrated over the last 10 years.

> which in effect subsidized smaller organizations at the expense of larger ones while simultaneously prolonging the time frame over which such smaller organizations can continue to ignore the global reality vs. allowing everyone the opportunity to obtain (costly) IPv4 space on a level playing field. 
> 
> 
> Even IPv6 will become more costly eventually as resource members switch from v4 to v6. How else will RIR's continue to exist. So lets not pretend like IPv4 is more costly than IPv6 save for the AfriNIC region where cost is low for motivation purposes.

When I talk about “costly” IPv4, I’m not talking about the RIR fees. That’s a drop in the bucket. I’m talking about the costs of CGN gateways, of technical support folks to answer irate phone calls about stuff that doesn’t work through CGN or NAT64 or whatever other band-aid and bailing wire contraption is next cobbled together as a form of IPv4 life support. I’m talking about the ever increasing cost of putting more and more subscribers into the same address space.

IPv6 completely eliminates those costs.

>> This way, we achieve forced IPv6 deployment. 
> 
> I don’t think forced is the right word here.
> 
> What I meant is that lack of available IPv4 is good news because IPv6 deployments will become a reality and new entrants will quickly greenfield IPv6 on the go which is why the so called Intra-RIR, inbound or outbound options to are bad for AfriNIC service region.

If IPv4 is really unavailable, then, yes, it helps.

However, what SL-BIS does is it creates this sort of meta-state where small organizations can still function with available IPv4 forcing large organizations to attempt to obtain additional address space through a transfer market. It’s kind of the worst of both worlds. All the pain of IPv4 runout without the IPv6 incentives it should provide.

> My point is that the costs of failing to deploy IPv6 are not entirely borne by the organizations that fail to deploy IPv6. Indeed mostly they are inflicted upon the organizations that have deployed it by those that have not. 
> 
> Expected effects of transitioning. 

Also known as the “toxic polluter economic model”.

Owen

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.afrinic.net/pipermail/rpd/attachments/20180607/38868423/attachment.html>