You run BGP over a 6to4 tunnel using 2002::/16 address space as the next hop, advertising your real PI/PA address space.
David -- http://dcp.dcptech.com > -----Original Message----- > From: [email protected] [mailto:ipv6- > [email protected]] On Behalf Of Tayeb > Meftah > Sent: Friday, May 10, 2013 9:48 AM > To: Ole Troan > Cc: Max Tulyev; [email protected] > Subject: Re: http://www.6assist.net/ - call for test > > Ole > How do bgp tunelling over 6to4 works? > > Envoyé de mon iPhone > > Le 10 mai 2013 à 15:02, Ole Troan <[email protected]> a écrit : > > > Max, > > > >> I mentioned Slovenia as we have a request for BGP-enabled tunnel from > >> Slovenia ;) So for some reasons people still want to use BGP-enabled > >> tunnels in real life, even in conuntries with well implemented native > IPv6. > >> > >> The second reason to use 6assist instead of regular TB it is not depend > >> of the actual load of tunnel server. If somebody download something > huge > >> through a tunnel broker server - the other people just share the tiny > >> rest of the bandwidth... > > > > if you go down this path, you could rather do BGP tunnelling. > > as in run BGP sessions between the peers, exchange native IPv6 prefixes, > but use a 6to4 > > next-hop. that achieves the mesh properties you are looking for with an > existing mechanism. > > of course someone will have to advertise a default route somewhere. > > > > cheers, > > Ole
