Hello, Le Mercredi 17 Mai 2006 16:09, vous avez écrit : > What is this configuration changed supposed to affect, exactly? It changes which prefix Windows expect from the server. > Does the MS Teredo server accept a client request for address with > this prefix? Not yet, or at least not at teredo.ipv6.microsoft.com. I couldn't get any infos as to whether and when this is going to change. > Is the global routing for it really working? The IANA prefix (2001:0::/32) is working; ITGate is advertising the prefix and running a working relay. The experimental prefix (3ffe:831f::/32) is not (anymore). > Is Teredo reliably and universally working with ANY configuration? No. No tunneling system works universally anyway, with the possible exception of DNS tunneling. Now, if you mean any NAT configuration, assuming there is no firewalling problems as well, then, no, Teredo as it is specified does not support symmetric NATs. There are undocumented-unspecified extensions in Windows Vista that would allow traversal of any NAT, provided either side of the packets exchange is behind no NAT or a friendly NAT. This would imply NAT traversal would at least work properly toward the whole IPv6 Internet (because Teredo relays are *NEVER* behind unfriendly NATs), and toward most Teredo clients behind SOHO NAT routers. Unfortunately, given this is not a specified extension, and that I'm fairly confident it's in explicit contradiction with the current specification, I have some reservations as to whether it's actually going to work as advertised. That is to say, if Teredo relays are running Vista, it might work fine, but not otherwise. I'd be happy to provide support for this in my own implementation, but I can't do so with zero concrete information. Also, the guys behind the NICI-Teredo implementation have seemingly developped their own extensions that, I assume, would allow Teredo to work in all cases. But their results is yet to be released: “Huang, S.-M., Wu, Q.-C., and Lin, Y.-B. Enhancing Teredo IPv6 Tunneling to Traverse the Symmetric NAT, Accepted and to appear in IEEE Communications Letters.” (quoted from http://liny.csie.nctu.edu.tw/) Regards, -- Rémi Denis-Courmont http://www.simphalempin.com/home/
Attachment:
pgpvxAJFC61KT.pgp
Description: PGP signature