India has a 63% IPv6 adoption while the USA 44%, Canada 34%, UK 33%, Korea 14% | Hacker News Reader

India has a 63% IPv6 adoption while the USA 44%, Canada 34%, UK 33%, Korea 14%

google.com

3 pointscuriousmindz6 comments

That seems odd. Does anyone know why?

Also, France and Germany are around 50%...

It's not odd at all. Reliance/Jio deployed a clean-build 4G network, and went dualstack. They also entered a price war which went gangbusters. Then the next big 3 providers decided to tool up. So now the top 4-5 ISps in India all "do" V6 and between them they have 80%+ of the Indian online experience.

Mobile is king in India. Low levels of Infra.

Korea (and China) dispute the measurement: they both say the external parties (Google, Akamai, APNIC (where i work) and others) mis-measure. I think the mis-measures agree so much, it can't be a mistake.

Thanks for the info.

> I think the mis-measures agree so much, it can't be a mistake.

What do you mean?

Edit: That they all independently measure and reach the same 14%?

It might be that Korea invested a lot, early on, to have great Internet coverage and speed (which is the case), but now they have a lot of hardware that is incomptable with IPv6. And it isn't worth it to replace all that just for that feature.

Yes, but also No: the point here is that S-K telecom say they have fantastic IPv6 penetration and market share, but we can't show it as IPv6 capable from probes in clients. There is a story about V6 in Korea which is about the early broadband adoption: they went with whitebox FTTP to the huge housing estates and the whitebox CPE can't do V6, nor can the in-ground switching fabric behind it. So, Korea overall will have to wait for the initial broadband investment to age out. (probably on the next 5-10 year window)

This problem, the mis-measure is probably about KR transit wierdness, and how we "see" viable IPv6 they may well have but how it exposes as end-to-end isn't good. Its a very odd market. Dacom/Boranet has done some de-peering at times, forcing domestic flows offshore, all kinds of wierdness in their market.

(yes, the point I was making is that 4 independent V6 deployment measures outside of the economy say the same figure for the economy, but domestic strategic planners are told by the ISPs they have far more IPv6 than we all see. Because we are independent but see the same approx 14%, I don't think we are making a mistake. Whats going on isn't clear to me)

> Does anyone know why?

Jio.[1][2]

[1] Leveraging IPv6 for Explosive growth – 306Mn+ Subs and counting (https://www.innog.net/wp-content/uploads/2019/07/01-Leveragi...)

[2] https://en.wikipedia.org/wiki/Jio

France is dominated by a small number of ISPs and went with a form of IPv6 to the home market which maybe put them in a bit of a dead-end (6rd) -which was invented in France. A very common French technolgy story, strong preference for the local solution (think Minitel) -But, they did get wide deployment.

Germany is dominated by DT (Deutsche Telekom) and what they do, moves the market. the UK is all about Sky (was News ltd, now Comcast, who are also behind a lot of the US domestic IPv6). Sky do remote management of the CPE and were able to upgrade their entire customer base (mostly) with remote upgrade version-by-version. Really high penetration. This caused Virgin and BT to have to think about how to compete for mindshare.

Reply on news.ycombinator.com