Re: [BEHAVE] Comments on draft-xli-behave-v4v6-prefix-00.txt

Xing Li <xing@cernet.edu.cn> Sun, 14 June 2009 00:21 UTC

Message-ID: <4A34428B.1030301@cernet.edu.cn>
Date: Sun, 14 Jun 2009 08:21:31 +0800
From: Xing Li <xing@cernet.edu.cn>
User-Agent: Thunderbird 2.0.0.21 (Windows/20090302)
MIME-Version: 1.0
To: Dave Thaler <dthaler@windows.microsoft.com>
References: <E4561B14EE2A3E4E9D478EBFB5416E1B0582A8@TK5EX14MBXW653.wingroup.windeploy.ntdev.microsoft.com>
In-Reply-To: <E4561B14EE2A3E4E9D478EBFB5416E1B0582A8@TK5EX14MBXW653.wingroup.windeploy.ntdev.microsoft.com>
Content-Type: text/plain; charset="UTF-8"; format="flowed"
Content-Transfer-Encoding: 8bit
Cc: "congxiao@cernet.edu.cn" <congxiao@cernet.edu.cn>, 'Behave WG' <behave@ietf.org>, Fred Baker <fred@cisco.com>
Subject: Re: [BEHAVE] Comments on draft-xli-behave-v4v6-prefix-00.txt
Precedence: list

Hi, Dave,

Thanks for providing very useful comments, which we'll consider very 
seriously and we'll make modifications and adjustments in the prefix 
document.

Concerning Section 6.1, here are the updates which reflect your comments.

Comment [DT64]: Disagree. See comments earlier, it’s up to the admin 
since both can work. It’s up to whether the admin wants to use multiple 
site prefixes (WKP and LIR both work) or just one (must use LIR).

Comment [DT65]: Disagree. Earlier in this document it already motivated 
why WKP is better the LIR for scenarios like “An IPv6 network to the 
IPv4 Internet”. Hence this appears to be the wrong recommendation for 
that scenario since there’s no justification for the LIR in this 
document for that scenario. Either fix the recommendation or add 
sufficient justification earlier.

Comment [DT66]: Disagree. The 6to4 prefix is intended to be advertised 
into the global IPv6 internet. A WKP is not. This is very different 
practice.

Please comment.

Regards,

xing

---------------------------------------------------------------------------------------------

6.1.  PREFIX Recommendation

   The PREFIX Recommendations are:

   o  In the case when different sites are using same IPv4 addresses
      (for example, [RFC1918] space), the LIR MUST be used.

   o  In the "an IPv4 network connecting to IPv6 Internet" scenario, the
      LIR MUST be used.

   o  In the stateless mode, it's up to whether the admin wants to use
      multiple site prefixes or just one.  If only one site prefix is
      used, it MUST be LIR.  If multiple site prefixes are used, the
      site prefix with global IPv6 connectivity MUST be LIR, while the
      site prefix without global IPv6 connectivity CAN be WKP.

   o  When the site admin uses LIR, all of the relevant protocols and
      software need to accommodate the ability to configure that LIR
      prefix.

   o  If for some reason, the admin cannot use LIR (e.g. in an isolated
      network), the admin should use this WKP allocated by IANA for this
      purpose, rather than pulling one out of thin air, or using a
      prefix allocated for a different purpose.

   o  When the WKP is used, it MUST not be propagated in native IPv6
      routing, to prevent pollution of the IPv6 routing table by
      elements of the IPv4 routing table.  Therefore, a site which also
      has a native IPv6 connection MUST NOT advertise its WKP routing
      prefix on that connection, and all native IPv6 network operators
      MUST filter out and discard any WKP routing prefix advertisements.




Dave Thaler 写道:
>
> Hi Xing Li,
>
> Thanks again for writing up this draft. I finally had time to type up my
>
> personal comments on the currently posted version of this document.
>
> Some of them repeat things others already said.
>
> -Dave Thaler
>

[BEHAVE] Comments on draft-xli-behave-v4v6-prefix… Dave Thaler
Re: [BEHAVE] Comments on draft-xli-behave-v4v6-pr… Xing Li