Mališa Vučinić <[email protected]> wrote: > I am not sure I follow. How come is it not the case that this well-known > host name is resolved to an IP address?
We don't need to map this name to an IP address, nor could we have a consistent mapping world-wide. The IP address is known to the Join Proxy (6LBR root or other JRC address provided to JP). The CoAP traffic is directed at the JRC by this process. The only reason we need the name at all is because we have to fill in the Uri-Host. Well.. I just re-read https://tools.ietf.org/html/rfc7252#section-5.10, and I realize that Uri-Host is not required, so actually we should just drop it entirely. > When JP receives a first DIO and learns the DODAG root’s IPv6, it stores > the IP address as the resolution of 6tisch.arpa. Same thing when it learns > the JRC’s IPv6 through the Join Response. If you want to think of it that way, sure, the JP is storing a mapping of 6tisch.arpa to that IPv6. I'm saying that *DNS* won't tell us the answer. > When a CoAP message to be proxied is received, and URI-Host that is in the > clear indicates 6tisch.arpa, the name is resolved to the previously stored > IP address and a new CoAP message is generated creating a new IPv6 packet. > Note that URI-Path is encrypted and therefore not available to the > proxy. > How is then 6tisch.arpa never associated with an IPv6 address? In your model, it could do that. I didn't think of it as working like that, but in my model all traffic that the JP receives via that channel goes to the JRC: it has a single purpose. In your model, the JP could proxy traffic to any place, I guess, but I didn't really want that to happen. -- ] Never tell me the odds! | ipv6 mesh networks [ ] Michael Richardson, Sandelman Software Works | network architect [ ] [email protected] http://www.sandelman.ca/ | ruby on rails [
signature.asc
Description: PGP signature
_______________________________________________ 6tisch mailing list [email protected] https://www.ietf.org/mailman/listinfo/6tisch
