> >If this tradeoff comes up with PMTUD makes sense when there is no > >tunneling, then I don't see why the same conclusion doesn't hold > >for the case of nested PMTUD using tunnels. > >So I'm looking forward to your performance analysis. > > i think i have answered this (partially) in other emails. > we are from very different assumption.
I'm still looking for the answer. What I'm trying to understand is why the "fragmentation considered harmful" arguments (see http://research.compaq.com/wrl/techreports/abstracts/87.3.html) don't apply to the tunnel. > then what is the IPv6 link MTU for tunnel interface, when you don't > have IPv4 PMTU information? (like for the very first packet to go out > from the path) You always have IPv4 PMTU information - if there isn't information from ICMP "too big" messages, then there is the interface MTU for the outgoing IPv4 interface to bootstrap the process. This is robust since the IPv4 path MTU can never exceed this interface MTU. Erik -------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
