[NCAP-Discuss] Current Status of the NCAP Project

[rubensk at nic.br]

> On 15 Nov 2022, at 05:08, Rod Rasmussen <rod at rodrasmussen.com> wrote:
> 
> Just to reset some folks’ understanding of IPv6 adoption, I recently was quite shocked by a presentation by Google at M3AAWG this fall on that very topic.  Google is seeing 40% of their traffic coming from native IPv6.  We should be reaching a tipping point soon where a majority of traffic (at least to Google) is over v6 and not from v4.

Our own share of credentialed users preferring IPv6 is 50%(measured at login time), and when it grows a bit more you will see press-releases from us celebrating that. But preferring IPv6 is very different from only supporting IPv6.

> 
> https://www.google.com/intl/en/ipv6/statistics.html
> 
> The growth rate has been steady and robust over the last 6-7 years, and what I had in my head of under 20% due to old numbers and thinking was completely shredded by actual data.  I’m going to be shelving my v6 jokes going forward...

We have some stickers with horse carriages calling IPv4 a legacy protocol, so we can joke about IPv4 right now... but what has been the growth rate of IPv6-only hosts ? That's the number to look for regarding collision mitigations.

> 
> This is largely driven by mobile (duh) but is highly geographical as well (see above website).  And yes the “server side” is still vastly different, but making assumptions about v6 based on the industry group-mindset we’ve had for years about v6 isn’t actually so wise.

Mobile IPv6 is almost always IPv4 + 464 XLAT. Mobile devices are frequently upgraded and that favored the adoption of newer technology. And yet they preserved compatibility with IPv4-only hosts by dual-stacking the host and the mobile operator border while using IPv6 for the transport network. 464 XLAT is possibly one of the cleverest technologies of the Internet Protocol world.
> 
> Just something to consider in discussions - we just don’t know how things are going to look 5-6 years from now based on assumptions made today.


Actually, ipv6.br <http://ipv6.br/> has been running for 14 years and most predictions of that time were on point on what would be the trigger factors for deployment, like adoption by large content providers and IPv4 space exhaustion.
We do some exercises now and then telling members of our network operators group to work with only IPv6 for a day, during our live events we once removed IPv4 from the conference Wi-Fi for the duration of a presentation. The end result is an Internet where you can search for content (Google Search is IPv6-capable for very long now) but can't click on almost anything. I suggest that experience for anyone believing an IPv6-only Internet is happening in their lifetimes.


Rubens

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mm.icann.org/pipermail/ncap-discuss/attachments/20221115/5e2338aa/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 529 bytes
Desc: Message signed with OpenPGP
URL: <https://mm.icann.org/pipermail/ncap-discuss/attachments/20221115/5e2338aa/signature.asc>