Re: [v6ops] Clarifications on draft-ietf-v6ops-rfc3316bis-00.txt

Excuse me quite late reply, inline please.

2012/11/18 jouni korhonen <jouni.nospam@gmail.com>

> Hi,
>
> On Nov 17, 2012, at 6:56 PM, Hui Deng wrote:
>
> > Hello, authors,
> >
> > in section 2.11, it says
> > "learning the DNS server  addresses from the link layer signaling can be
> cumbersome when the MT
> >    and the TE are separated using other techniques than PPP interface"
> > Can you be more specific about what else could be used? and what can be
> cumbersome?
>
> Like proprietary APIs or USB cdc etc between the MT and the TE. There
> might not be "easy" or standard way to let the TE stack to learn DNS
> information that the network sent in NAS signaling. PPP had one. Learning
> the DNS information using IP protocols between the TE stack and the gateway
> just makes the MT implementation easier and avoids OS specific hacks there.
>
I thought propriatary way also works, but if you think so, please add some
text in the current draft.

>
> > for privacy consideration,
> > can you help to write some analysis text about the relationship with
> "always on capability in 3GPP"
>
> Do you mean analysis of a cellular host having the same prefix in use for
> a long periods of time?
>

What I mean is in the case of LTE, it will keep one IP address or IPv6
prefix to allow its always on which keep PDP context.

>
> > Does SIPTO/LIPA has some influence on this document?
>
> I am not sure yet. Maybe a recommendation to use DNA might be useful here?
>

In the case of SIPTO/LIPA, UE will has two PDP connections and two IP
addresses assigned. does it make some level difference?

thanks for your discussion

-Hui

>
> - Jouni
>
>
>
> >
> > thanks
> >
> > -Hui
> > _______________________________________________
> > v6ops mailing list
> > v6ops@ietf.org
> > https://www.ietf.org/mailman/listinfo/v6ops
>
>