In your letter dated Wed, 20 Apr 2011 15:01:05 +1200 you wrote: >In this situation a 6to4 relay (like any other tunnel end point) >should behave according to section 3.2 of RFC 4213. That's quite >a complicated section and I suppose there may be buggy >implementations, even in the absence of ICMP filtering.
That illustrates what I'm trying to say. Sending packets with an MTU of 1280 is always safe. So to be able to send packets with an MTU of around 1500, we have a lot of complated machinery that has been shown to be very fragile, both in IPv4 and IPv6. And all we get from that is a rather modest efficiency gain. At the cost of breaking things completely when it goes wrong. So I would suggest that we strongly recommend to host operating system implementors and content providers that they make sure that a sending system's MTU is quickly reduced in the presence of an MTU blackhole. -------------------------------------------------------------------- IETF IPv6 working group mailing list [email protected] Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
