You're making a host of assumptions here. One of them is that even though the info is requested per-host, it exists as per-site.
no, I'm not making that assumption. the only assumption I'm making is that the mappings from id->loc are the same for every host, so that they can be cached if they are used by multiple hosts.
Yes, that was what I was saying.
I am fairly convinced that id-to-loc mapping needs to be more like mobile-ip where you send a packet to the id and get a redirect to the loc if the network can't route to the id,
But what if the network silently (or seemingly silently, re ICMP filtering) drops your packet?
if the network does that, it's broken.
Surprise: the network IS broken. Operators have turned to large scale ICMP filtering to be able to keep limping along during the recent worm season.
we can design protocols to
tolerate transient failures; but we cannot design protocols that work no
matter what arbitrary filtering the network imposes.
I agree in principle but in practice a protocol that depends on the random kindness of remote routers won't fly.
the network has a responsibility to make a best effort to deliver packets to their destination intact. this is nothing new.
Best effort may not be enough, not new either.
(note that I never said that the protocol would use ICMP for signalling such things.)
Neither did I, I only cited ICMP filtering as an example. But ICMP is the protocol of choice for control messages.
-------------------------------------------------------------------- IETF IPv6 working group mailing list [EMAIL PROTECTED] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
