Hi Bjorn, > -----Original Message----- > From: Bjørn Mork [mailto:bj...@mork.no] > Sent: Friday, January 24, 2014 4:10 AM > To: Templin, Fred L > Cc: ipv6-ops@lists.cluenet.de > Subject: Re: FW: I-D Action: draft-foo-v6ops-6rdmtu-01.txt > > "Templin, Fred L" <fred.l.temp...@boeing.com> writes: > > > Concerns for > > operational issues with both IPv4 and IPv6 Path MTU Discovery point > > to the possibility of MTU-related black holes when a packet is > > dropped due to an MTU restriction. > > So the "fix" is to just let PMTUD die for IPv6 too? Or should blackholing > anyone dropping ICMPv6 Packet Too Big errors be considered a feature > instead?
No; not let it die. But rather, let the network be an *advisory* source of PMTU information and let the end systems themselves be the *authoritative* source. > Forget about "World IPv6 Day". We need a "World Path MTU Day", where > everyone running a public service of some kind lowers the MTU on links > leading to their servers to a random number between 1280 and 1500. No, we want to get at least 1500 everywhere. And, allow for larger sizes when they are available. Thanks - Fred fred.l.temp...@boeing.com > Bjørn