Re: [v6ops] The need for local-ipv4 socket transition solutions -- NAT64/DNS64 remains insufficient

[Xing Li <[e-mail]>]

Ross Chandler 写道:
>> On 7 Apr 2015, at 06:23, Xing Li <[e-mail]> wrote:
>>     
>>> Once IPv6-only with double IPv4/IPv6 translation is available the next stage to reach is IPv6-only. IPv4 literals will be with us for as long as IPv4 is.  
>>>       
>> If all the users on the Internet can access both the IPv4 (via double and single translation) and IPv6 (native) contents, who cares if the IPv4 literals will coexist with IPv4 and IPv6 for a long time?
>>     
>
> I think for 3GPP especially people care more about the IPv4 literals & IPv4 only apps issue as that necessitates IPv4+IPv6 on the PDP/PDN connection for Internet access. Having 464xlat stops this issue hindering adoption of IPv6. With the local-ipv4 socket transition solution efforts to fix IPv4 literals and IPv4 only apps can proceed in parallel with IPv6 deployment instead of impeding it.
>
> In fixed Internet, not having a local-ipv4 socket translation solution isn’t as bad a hinderance as the end hosts sit behind a CPE providing dual-stack on the LAN side. 
>   

In order not to downgrade the users experience, the choices are:
(1) Dual-stack: dual-stack network interface and dual-stack local sockets
(2) 464xlat:  IPv6-only network interface and dual-stack local sockets

I believe (2) provides a better chance to move to IPv6-only then (1), no 
matter with or without a CPE. Can we convince all the users turn off the 
IPv4 socket in their OS before the *whole* Internet fixes the IPv4 
literals & IPv4 only apps problem, even their ISP has IPv6 and NAT64 
services? Frankly speaking, it is very difficult, if not impossible.

Regards,

xing




> Ross
>
>
>
>