Hi Joel, Thanks for the clarification. Fully agree with the 4 types.
Two additional comments: 1. From implementation view, the 2nd and the 3rd are more close to the 4th because there are binding table lookup procedure in data plane, while the 1st depends on algorithmic address calculation essentially. Some may argue that there are mutliple rules even in the 1st case, but that's a specific feature in IPv4-over-IPv6 scenario with address sharing demand, only because we want to include port set information into IPv6 address beside the IPv4 address and there may be not enough bits in IPv6 address. Generally this is not a necessary feature for pure stateless solution.(take 6RD and SIIT as example) 2. In this specific context, I quote what is the defined in the stateless motivation draft(http://datatracker.ietf.org/doc/draft-ietf-softwire-stateless-4v6-motivation/): " Stateless 4/6 solution (or stateless solution in short): denotes a solution which does not require any PER-USER state (see Section 2.3 of [RFC1958]) to be maintained by any IP address sharing function in the Service Provider's network. This category of solutions assumes a DEPENDENCY between an IPv6 prefix and IPv4 address." On Mon, Jun 25, 2012 at 9:39 PM, Joel M. Halpern <[email protected]> wrote: > This appears to mix different kinds of state. > In particular, with regard to state in network resident boxes, there can be: > General state covering all subscribers; > Per subscriber state provisioned state; > Per subscriber dynamic, but not flow specific state; > per flow state. > > All are state. > The first is clearly permitted. > The fourth is clearly not permitted in a "stateless" solution. > Some descriptions of "stateless" have allowed for the middle two, while > others have prohibited it. I have no opinion on what the WG "agreement" is > on the scoping. But we need to be careful about what we mean about "state". > > Yours, > Joel > > > On 6/25/2012 2:51 AM, Qi Sun wrote: >> >> Hi Satoru, >> In MAP 1:1 mode, if there are 10000000 subscribers, there would be >> 10000000 MAP domains which a BR has to manage. I think that will create >> a huge mapping table on the BR, which is called 'state' that stateful >> solutions deal with. >> Best Regards! >> Qi Sun > > _______________________________________________ > Softwires mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/softwires _______________________________________________ Softwires mailing list [email protected] https://www.ietf.org/mailman/listinfo/softwires
