In message <[email protected]>, =?iso-8859-1?Q?R=E9m i_Despr=E9s?= writes: > > Le 3 mars 2011 =E0 20:48, Brian E Carpenter a =E9crit : > > ... > > Of course. PI has the cost of exploding the BGP4 table. > > NPTv6 has the cost of destroying address transparency. > > = > > > Since SHIM6 has neither of these costs, are you surprised > > that I prefer it? > > SHIM6 is a solution, but AFAIK it needs that hosts know: > - their global addresses > - means to force outgoing paths that go via the right CPE's (an ingress-fil= > tering constraint) =
Having a per announced prefix default route would definitely help here, for SLAAC at least. If a router annouces a prefix then it is a candidate for the default for packets sourced from the addresses within that prefix. > In a site using ULAs and having several links, a complement is therefore ne= > eded. = > > SAM is an approach for this (ref. draft-despres-softwire-sam-01 - sec. 3.3)= > , and AFAIK the only documented one so far. > Would you agree on this? > > RD > > > > > _______________________________________________ > nat66 mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/nat66 -- Mark Andrews, ISC 1 Seymour St., Dundas Valley, NSW 2117, Australia PHONE: +61 2 9871 4742 INTERNET: [email protected] _______________________________________________ nat66 mailing list [email protected] https://www.ietf.org/mailman/listinfo/nat66
