Re: [v6ops] v6ops-host-addr-availability: A Little Pushback

[Dan Drown <[e-mail]>]

Quoting Lorenzo Colitti <[e-mail]>:
:
> As someone who has worked on the ePDG implementation on Android devices, I
> can assure you that the architecture that's used to ensure that the
> baseband processor and the application processor could share an IP address
> is pretty contorted - you might even say nightmarish.
:
>> <bhs> NAT66 is not something that hosts inside enterprise networks need to
>> be worried about.
>>
>
> Why? I know lots of users of my enterprise network that run VMs, run
> Android / iOS emulators on their development machines, use ePDG services,
> etc. With only one /128 per host, you can'd to that without NAT66.

Because I didn't know what ePDG was before this, I had to look it up.   
Below is my summary of it, anyone who understands it better, please  
feel free to correct it.

ePDG is a secure tunnel back to your cell phone operator, and wifi  
calling is one application that can run over it.  It runs on the  
baseband processor rather than the application processor of the phone.  
  There are two different operating systems running on these  
processors, and they need to share connectivity to the outside world.   
With v4, I assume this is achived with NAT44 when connected to a wifi  
AP.

Wifi calling is an application that's on millions of phones already.   
I don't know if ePDG is a popular implementation or if there are any  
competing implementations.

With v6, the design of assigning a dedicated v6 address (from a prefix  
or a pool) to the baseband processor would make sense for this  
application, as it makes the phone's internal network easier to  
understand.  The alternatives would be to do NAT66 on the application  
processor or not have v6 on the baseband processor at all.