[afnog] [rpd] Re: A typical case of abuse of our resources!!!

[Owen DeLong]

On Sep 24, 2014, at 7:11 AM, Jackson Muthili <jacksonmuthi at gmail.com> wrote:

> On Wed, Sep 24, 2014 at 2:34 AM, Owen DeLong <owen at delong.com> wrote:
>> Asking AfriNIC staff to implement the spirit of the policy rather than
>> taking care to write what we really want as a community into the letter of
>> the policy is a very dangerous request.
>> 
>> If you had a loosely written policy to a dozen people and ask each of them
>> to explain the spirit of the policy, it is unlikely that more than one or
>> two of them (at best) will come particularly close to the original author’s
>> stated intent for the policy.
>> 
>> More importantly, you will probably have at least 12 different (and often
>> some of them radically different) ideas as to the core intent of the policy.
>> 
>> I agree that the allocation is suspicious, but if the postmasters say that
>> the justification supplied meets policy, then I believe we must trust them
>> absent any strong evidence of fraud. Share suspicion alone is not
>> sufficient.
>> 
>> I suspect that the domains are being used to host VPNs and are likely
>> serving end-users outside the region via these VPNs. Currently, there’s
>> nothing in AfriNIC policy to prohibit that. I leave it as an exercise for
>> other members of this list to decide whether there should be policy to
>> prohibit such allocations.
>> 
>> I will say this… Regardless of how we hand out the remaining IPv4 addresses,
>> the simple reality is that IPv4 will soon be much less important than having
>> an IPv6 deployment and I believe that the community would be better served
>> by aggressively implementing IPv6 than by handwringing over the mechanism by
>> which the limited IPv4 free pool is drained.
> 
> While you keep saying this you did not say some importants facts like
> (1) Ipv4 needed during transition (2) 98% of internet still on ipv4
> despite that you think ipv4 is not anymore important (3) the develope
> or 1st world economy (USA include) have not adopt ipv6. I wonder why?

1. At this point, there are few dependencies on IPv4 for transition. There are
dependencies on IPv4 for maintaining connectivity to content which has not
yet adopted IPv6.

2. Depending on how you measure this, one could argue that much closer to
100% of the internet remains on IPv4. However, that’s irrelevant, as all dual
stacked hosts are “still on IPv4”. The fraction of the internet that is not yet
on IPv6 or does not yet have IPv6 service available is much smaller. It’s
actually closer to 60% at this point. The fraction of major content providers
that have not yet deployed IPv6 is approximately 84%.

(You can see this for yourself at http://ww.delong.com/ipv6_alexa500.html
which is updated daily at 0000 UTC).

3. Actually Comcast (the largest residential ISP in the US) has made IPv6
available to the vast majority of their residential users. Google, Facebook,
Youtube, Yahoo, LinkedIn, Netflix, and many others have made their content
available on IPv6. The first world has significantly deployed IPv6 and more
is being deployed every day.

At the present rate of IPv4 and  IPv6 growth, there will be more internet
users on IPv6 than on IPv4 in approximately 1.5 years. This assumes that
the current IPv4 growth rate is sustainable (it isn’t) and that the IPv6 rate
remains constant (it actually appears to be accelerating).

If IPv4 does not maintain its growth rate and IPv6 continues to accelerate,
likely that will be somewhere more like mid-2015.

> What I try to say here is ipv4 is still very important and please dont
> discourage developing economy from levereaging so far its pool larger
> than rest of the world because it seem that internet may still run on
> Ipv4 for long time.

You’re welcome to leverage whatever you wish. However, I point out to you
that your IPv4 inventory has a pretty limited shelf life at this point. I never
said it was not important today. I said it would soon be much less important
than it is today and I will stand by that statement.

I do not believe for one second that the internet can actually run on IPv4 for
a long time by any definition of long time > 5 years. It simply isn’t sustainable
due to a number of constraints which include routing table size, churn rates,
routing engine capabilities, TCAM limitations, and the utter destruction of
capabilities that comes with NAT and is exacerbated by CGN, not to mention
the security nightmare that is created by CGN in addition to the interesting
denial of service nature of the intersection of IP reputation systems (widely
used at the present time to make IPv4 usable) and CGN.

Owen