[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: translator friendly DNS Proxy

Sorry, I forget to respond to another question.
I think there are not big differences between my+masahito's proposal and Bagnulo's. There are some minor differences, like the method how to identify the synthetic address. Even this, the concept is very close to snapt, since both proposed to 
be recognize on around application layer.

Actually, the behavior is what I am insisting on my two proposals.
http://www.ietf.org/internet-drafts/draft-miyata-v6ops-trans-approach-01.txt
http://www.ietf.org/internet-drafts/draft-miyata-v6ops-snatpt-00.txt
And my presentation in 71st IETF. The presentation was done by Fred on behalf of me.
I and Masahito know that NAT-PT translation engine works with totd like dns proxy. We have the implementations for more than 5 years. The behavior is very popular among translator guys. It was introduced with TRT around 2000, as draft-endo-v6ops-dnsproxy-00.txt 
mentioned. ;-)
I would respectfully ask Bagnulo, what is the biggest difference between us. 


Dan, I agree with your modification request. Thanks a lot!

Kindly Regards,

....miyata
On 2008/09/05, at 1:15, <H.Miyata@jp.yokogawa.com> <H.Miyata@jp.yokogawa.com > wrote: 


Oops,

Sorry, I would like to maintain it available.
Thanks for your comments.

Thanks,

.....miyata

________________________________________
差出人: Dan Wing [dwing@cisco.com]
送信日時: 2008年9月4日 7:39
宛先: Endou, Masahito (Masahito.Endou@jp.yokogawa.com); v6ops@ops. ietf.org 
CC: behave@ietf.org; Miyata, Hiroshi (H.Miyata@jp.yokogawa.com)
件名: RE: translator friendly DNS Proxy


-----Original Message-----
From: Masahito.Endou@jp.yokogawa.com
[mailto:Masahito.Endou@jp.yokogawa.com]
Sent: Thursday, August 07, 2008 8:02 PM
To: dwing@cisco.com; v6ops@ops.ietf.org
Cc: behave@ietf.org; H.Miyata@jp.yokogawa.com
Subject: RE: translator friendly DNS Proxy

Hi, Dan


Hi, all

I submitted this document.

In this document, I proposed DNS proxy that is separated

from NAT-PT.

This document describes about relationship DNS proxy and

sNATPT[1].

I think that this DNS proxy can collabolate with other translation
technologies.

I know that this working group isn't appropriate to discuss

such kind

items, but I informed this document bacause many people that are
interested in subscribed this mailing list.
And, if there is more suitable working group to discuss it, please
tell me that.


The Behave working group would be best,
<http://www.ietf.org/html.charters/behave-charter.html>.  I
just forwarded your announcement to the Behave mailing list.


Thanks for your information.
I just subscribed :)


[1] sNATPT: Miyata, H., "Simplified Network Address Translation -
Protocol Translation"
                  draft-miyata-v6ops-snatpt-00


I read your draft, and it seems like the Translator Interface
is the key -- it is the interface between the DNS Proxy
(which synthesizes AAAA records) you describe, and the NAT-PT
itself (which NATs between v6 and v4).

Do I understand correctly?


Yes.
But actual key point is that the translator interface can
separate DNS Proxy from NAT-PT.

If a translator has NAPT-PT rule for IPv6 to IPv4
translation, the translator interface is not required.
In this case, a translator maps a lot of IPv6 source
addresses to one IPv4 source address,
so DNS Proxy doesn't need to care an IPv4 source address.


(Sorry for my delay.)
I noticed that draft-miyata-v6ops-snatpt-00 has expired. Did you want to continue discussing it? We have not included it in the upcoming comparison 
document.


In the draft-miyata-v6ops-snatpt-00:
In the IPv6->IPv4 direction, the separation of DNS rewriting ("DNS- ALG") from NATing seems very similar to draft-bagnulo-behave-nat64-00. Can you describe 
the differences, if any, between draft-miyata-v6ops-snatpt-00 and
draft-bagnulo-behave-nat64-00?
In the IPv4->IPv6 direction, Section 5.2 of draft-miyata-v6ops- snatpt-00 describes that both static v4->v6 mapping and dynamic v4->v6 mapping can be supported. The text appears to go on to describe dynamic mapping. I believe we would only need static mapping from IPv4->IPv6, as described very briefly 
in Section 6 of draft-miyata-v6ops-snatpt-00.

-d


best regards,

// masaxmasa


-d



Thanks your comments.
Best Regards,

--------------------------
A new version of I-D, draft-endo-v6ops-dnsproxy-00.txt has been
successfuly submitted by Masahito Endo and posted to the IETF
repository.

Filename:        draft-endo-v6ops-dnsproxy
Revision:        00
Title:           Translator Friendly DNS Proxy
Creation_date:   2008-08-07
WG ID:           Independent Submission
Number_of_pages: 24

Abstract:
This document describes the DNS Proxy that is separated

from NAT-PT

[RFC2766].  NAT-PT was designed to work with DNS Application Level
Gateway.  However [RFC4966] pointed out DNS related issues, and
[RFC2766] was changed to historical state.  This document

attempts to

DNS Proxy specification, removing dependency on NAT-PT as well as
resolving problems pointed in [RFC4966].

// masaxmasa








_______________________________________________
Behave mailing list
Behave@ietf.org
https://www.ietf.org/mailman/listinfo/behave