Med, >From protocol level, the difference between public 4over6 and lightweight 4over6(b4-translated-ds-lite) lies in port-set support. The extra efforts of lw 4over6 are as follows: (1) port set support in DHCP provisioning; (2) NAT on the initiator side.(whose address pool is not a full address but only a port set) (3) port-set supporting in the cocentrator's binding table.
While we may cover public 4over6 by lightweight 4over6 with a special port set format (2^16 size) for (3), (1) and (2) brings quite significant changes to the intiator side. If I'm only a pb 4over6 initiator, more typically a host initiator, all the complexity needed is to plant a CRA process on the host, which is basically an IPv4 & IPv6 socket function, to support DHCPv4-over-IPv6. The rest is already there: we don't need to modify DHCP client, and IPv4-in-IPv6 tunnel is already supported in today's OS. No NAT is needed in host case, and full e2e transparency is guaranteed. If we support this by lw 4over6, we implemented extra complexity which is not needed at all by the initiator. We have deployement scenarios which probably don't require address sharing, such as CERNET, and I guess maybe the ISPs in USA also do not have an IPv4 address shortage problem? So, aside from the fact that the pb 4over6 draft starts earlier and its status has been a step furher, this is a problem of choice: do we want two clean, simple mechanisms, or one mechanism trying to be compatible with both. On Thu, Jun 7, 2012 at 9:11 PM, <[email protected]> wrote: > Dear all, > > I agree with Reinaldo. > > IMHO it makes sense to merge the two documents: either > draft-ietf-softwire-public-4over6 be extended to cover > draft-cui-softwire-b4-translated-ds-lite or add one or two sentences to > draft-cui-softwire-b4-translated-ds-lite to mention a non-shared IPv4 address > may be assigned. > > Doing so would help to rationalize the solution space and associated > documents. We have the following main flavours: > > (1) Full stateful mode: DS-Lite > (2) Full stateless mode: MAP > (3) Per-customer state/binding mode: lw4o6 > (draft-cui-softwire-b4-translated-ds-lite) > > All the three modes must support the ability to assign a full IPv4 address. > > Cheers, > Med > >>-----Message d'origine----- >>De : [email protected] >>[mailto:[email protected]] De la part de Reinaldo Penno >>Envoyé : lundi 28 mai 2012 07:53 >>À : Sheng Jiang; Yong Cui; [email protected] >>Objet : Re: [Softwires] WG last call on >>draft-ietf-softwire-public-4over6-01 >> >>-1 >> >>In which significant way this document is different from >>http://tools.ietf.org/html/draft-cui-softwire-b4-translated-ds- >>lite-06 ? >> >>We can insert one paragraph in the above draft and allow >>public IPs since >>NAT is optional. The two documents even use DHCPv4ov6 as provisioning. >> >> >> >>On 5/27/12 6:32 PM, "Sheng Jiang" <[email protected]> wrote: >> >>>The document looks mature for being advanced. >>> >>>Sheng Jiang >>> >>>> -----Original Message----- >>>> From: [email protected] >>[mailto:[email protected]] On >>>> Behalf Of Yong Cui >>>> Sent: Sunday, May 27, 2012 10:31 PM >>>> To: [email protected] >>>> Cc: Yong Cui >>>> Subject: [Softwires] WG last call on >>draft-ietf-softwire-public-4over6- >>>> 01 >>>> >>>> Hi folks, >>>> >>>> This is a wg last call on draft-ietf-softwire-public-4over6-01. >>>> http://datatracker.ietf.org/doc/draft-ietf-softwire-public-4over6/ >>>> >>>> As usual, please send editorial comments to the authors and >>>> substantive comments to the mailing list. >>>> >>>> This wg last call will end on 2012 June 10 at 12pm EDT. >>>> >>>> >>>> Yong & Alain >>>> >>>> >>>> >>>> _______________________________________________ >>>> Softwires mailing list >>>> [email protected] >>>> https://www.ietf.org/mailman/listinfo/softwires >>>_______________________________________________ >>>Softwires mailing list >>>[email protected] >>>https://www.ietf.org/mailman/listinfo/softwires >> >> >>_______________________________________________ >>Softwires mailing list >>[email protected] >>https://www.ietf.org/mailman/listinfo/softwires >> > _______________________________________________ > Softwires mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/softwires _______________________________________________ Softwires mailing list [email protected] https://www.ietf.org/mailman/listinfo/softwires
