Re: How to differentiate traffic in NAT-PT box?

[Soohong Daniel Park <soohong.park@samsung.com>]

Withour DNS-ALG, we can live with NAT-PT as well.
http://www.watersprings.org/pub/id/draft-daniel-natpt-bis-01.txt 

Don't say "Killing NAT-PT". It seems a harmful opinion against
NAT-PT developer, although they are still weak status. Just saying 
"Clarifying NAT-PT". No technology can be a perfect level. 

Daniel (Soohong Daniel Park)
Mobile Convergence Laboratory, SAMSUNG Electronics.

----- Original Message ----- 
From: "Brian E Carpenter" <brc@zurich.ibm.com>
To: "haofeng Zhang" <hfzhang.cn@gmail.com>
Cc: <v6ops@ops.ietf.org>
Sent: Friday, February 10, 2006 9:03 PM
Subject: Re: How to differentiate traffic in NAT-PT box?


> One of the many reasons for killing NAT-PT.
> 
>     Brian
> 
> haofeng Zhang wrote:
>> Dear all,
>> 
>> I have a problem regarding to NAT-PT and ALG.
>> In the RFC2766(NAT-PT spec), nat-pt box cooperates with ALG to deal with
>> some application carrying IP addresses in payload. But for the operation of
>> nat-box box, how can the box differentiate which kind of traffic needs to be
>> processed by ALG and others needn't? Or the nat-pt box just send all traffic
>> to ALG module and leave the differentiation problem to ALG?
>> I don't find the clarification in the spec. So if answer is the latter, does
>> it means a big waste of device resource? After all, usually only the
>> signaling traffic needs to be processed by ALG.
>> So any kind of help is appreciated.
>> 
>> --
>> Best regards,
>> Zhang haofeng
>> 
> 
> 
> 
>