Hi Yiu,
When the WAN is IPv4-only, "CPE" can automatically get prefixes with 6to4 or
6RD to be used locally (just by learning 6RD ISP prefix or using 6to4 WKP).
When the WAN is dual-stack, by current specifications, "CPE" would have to use
DHCPv6 for obtaining additional IPv6 prefixes.
Now the "CPE" could use 6to4 or 6RD in the DS WAN case as well, but then it
would have to encapsulate.
I am wondering if it would make sense to have automatic prefix delegation, but
without need to encapsulate.
In the setups where CPE is behind dual-stack point-to-point link from gateway
(GGSN, PDN GW, Home Agent, LMA, ASN GW, BRAS(?)..), the gateway would know
where to route as it has assigned the "CPE"'s IPv4 address, which the "CPE"
would use in constructing its automatically delegated prefixes.
Best regards,
Teemu
>-----Original Message-----
>From: ext Lee, Yiu [mailto:[email protected]]
>Sent: 28 January, 2010 17:26
>To: Savolainen Teemu (Nokia-D-MSW/Tampere); [email protected]
>Subject: Re: [Softwires] 6RD prefix delegation without encapsulation?
>
>Hi Teemu,
>
>In 6rd design, we assume that the WAN network is IPv4-only. It is ok
>that
>the CPE WAN port can be dual-stack, but it will never get v6
>connectivity
>because the network is IPv4-only. So, what is the use case of your
>question?
>Can you explain it in a little more details?
>
>Thanks,
>Yiu
>
>
>On 1/21/10 5:19 PM, "[email protected]"
><[email protected]> wrote:
>
>> Hi,
>>
>> draft-ietf-softwire-ipv6-6rd-03 describes how prefixes are
>automatically
>> delegated.
>>
>> It also describes IPv6 in IPv4 encapsulation.
>>
>> However, if CPE has native dual-stack WAN connection, with BR locating
>at the
>> first hop (say point-to-point connection), what is the problem in CPE
>> obtaining these automatically delegated prefixes as described in 6RD
>(i.e.
>> based on CPE's IPv4 address), but without need to encapsulate IPv6
>packets in
>> IPv4?
>>
>> Thanks,
>>
>> Teemu
>> _______________________________________________
>> Softwires mailing list
>> [email protected]
>> https://www.ietf.org/mailman/listinfo/softwires
_______________________________________________
Softwires mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/softwires
