Hi Woj, Thanks for the editing! I'm fine with the description for lw4over6. As for the map part, I think it's not that accurate.
Here is another proposal, based on yours: Lightweight 4over6 is a solution designed specifically for complete independence between IPv6 subnet prefix and IPv4 address with or without IPv4 address sharing. This is accomplished by maintaining state for each softwire (per-subscriber state) in the lwAFTR and a hub-and-spoke forwarding architecture. [I-D.ietf-softwire-map] also offers a means for reducing the amount of such state using address mapping rules. However, this introduces a algorithmic relationship between the IPv6 subnet prefix and IPv4 address. This relationship also allows the option of direct meshed connectivity between users. The reasons are: 1) "same capabilities" This seems a marketing wording, like "optimizing". lw4over6 architecture is applicable for other provisioning mechanisms, like DHCPv4ov6/PCP, which adds flexibility to lw4over6. Potential flexibility would make them have different capabilities. So I think the word should be removed and keep this text purely technical. IMO, this also follows the WG's guidance during the Softwire session. 2) "algorithmic" It is the designed feature of MAP to use algorithmic v4-v6 mapping to achieve stateless. No matter what mode MAP is running, the algorithm is still there, but with different value of parameters. A vendor has to implement the algorithm if he wants to use MAP. So it is accurate to keep the algorithmic there, and it reflects the characteristics of MAP. Best Regards, Qi On 2014-3-17, at 下午5:11, Wojciech Dec wrote: > Hi Ian, > > here's an updated proposal. Besides lining up the text on the prefix > independence, I also edited out some redundant text... > > Lightweight 4over6 is a solution designed specifically for complete > independence between IPv6 subnet prefix and IPv4 address with or without IPv4 > address sharing. > This is accomplished by maintaining state for each softwire (per-subscriber > state) in the lwAFTR and a hub-and-spoke forwarding architecture. > [I-D.ietf-softwire-map] offers the same capabilities along with means for > reducing the amount of such state using address mappings rules, which however > introduce a relation between the IPv6 subnet prefix and IPv4 address. This > relation also allows optional meshed connectivity between users. > > Cheers, > Wojciech. > > > > On 13 March 2014 09:00, Ian Farrer <[email protected]> wrote: > Hi Woj, > > Can we resolve this with the following wording change? > > Old: > Lightweight 4over6 provides a solution with complete independence of IPv4 and > IPv6 addressing (i.e., the IPv6 prefix does not embed an IPv4 address and/or > port set). > > New: > Lightweight 4over6 is a solution designed specifically for complete > independence between IPv6 subnet prefix and IPv4 address with/without CPE v4 > address sharing. > > The complete paragraph with the proposed change would then read: > > Lightweight 4over6 is a solution designed specifically for complete > independence between IPv6 subnet prefix and IPv4 address with/without CPE v4 > address sharing. > This is accomplished by maintaining state for each softwire (per-subscriber > state) in the lwAFTR and using a hub-and-spoke architecture whereby all > traffic traverse > the lwAFTR. [I-D.ietf-softwire-map] offers a means for reducing the amount of > such state by using algorithmic IPv4 to IPv6 address mappings to create > aggregate rules. > This also gives the option of direct, meshed IPv4 connectivity between > subscribers. > > If not, please can you suggest an alternative wording? > > Cheers, > Ian > > > > > On 12 Mar 2014, at 13:00, Wojciech Dec <[email protected]> wrote: > >> Let me try again: >> Basic fact: IPv6 address is 128 bits. >> >> If you put in the IPv4 address in the IID, then it becomes part of the 128 >> bit IPv6 address. Claiming that there is complete independence (no linkage) >> between the IPv4 addresses and IPv6 address is then not correct. If you see >> it another way, then it would be interesting to understand why. >> What I think Qi was/is trying to say is that there is prefix independence, >> with prefix being the part of the IPv6 address less the IID. That I agree >> with. It is the same in lw46 and MAP-E >> >> Side note: The "MAP-E" algorithm derives a PSID and port ranges. That's it. >> This PSID, along with trailing bits of the IPv4 address *may* get inserted >> in MAP-E into the non IID IPv6 prefix and this is unique to MAP-E. >> >> Hope it helps. >> >> >> On 12 March 2014 12:41, Lee, Yiu <[email protected]> wrote: >> I don’t understand his point. Let’s put the 1:1 aside, MAP-E requires IPv4 >> rule to algorithmically build the CE IPv6 prefix. In lw4o6 Section 5.1, we >> simple put the v4 in the IID. Isn’t it obviously there is no v4/v6 >> dependency? What Woj tries to argue? I lost. Can somebody explain to me >> please? >> >> From: Wojciech Dec <[email protected]> >> Date: Wednesday, March 12, 2014 at 6:48 AM >> >> To: Qi Sun <[email protected]> >> Cc: Softwires-wg WG <[email protected]> >> Subject: Re: [Softwires] Proposed text that describes lw4o6 and map-e >> >> Not on reading section 5.1 of : >> http://tools.ietf.org/html/draft-ietf-softwire-lw4over6-07 >> Where text like "The /128 prefix is then constructed in the same manner as >> [I-D.ietf-softwire-map]" is present. >> >> I do agree that there is top level prefix independence. >> >> Regards, >> Wojciech. >> >> _______________________________________________ >> 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
