Re: The IPv4 Internet MTU

[Magnus Westerlund <magnus.westerlund@ericsson.com>]

MTU's are real bastard. Especially for UDP applications that rarely has
built in any support for handling MTU discovery if the ICMP doesn't
work. Which it clearly do not do. The worst is that each
application/protocol needs to support MTU handling by itself.

At least for TCP we at least have a solution that can be gradually
deployed in RFC 4821 "Packetization Layer Path MTU Discovery" that
doesn't rely on ICMP. Simply getting this implemented in a majority of
the TCP stacks would be really good.

However, I think we need to strive towards getting applications in line
and support MTU discovery. It is silly for high-bit rate applications to
send all these small packet instead of at least descent sized packets
when possible.

Rémi Denis-Courmont skrev:

> Maybe Fred was hinting that Teredo could have defined its
> 
> fragmentation layer between UDP and IPv6. But while this would
> 
> work-around on-path fragmentation it would bring a naughty congestion
> 
> control issue - one simply cannot send 40 datagrams of 1500 bytes each
> 
> with no pacing. And we don't want to have to do congestion control,
> 
> right?
>

Exactly, and you also need to define a reliability mechanism so you can
retransmit fragments that got lost so that you don't get the packets to
experience only (1-packet_loss_probability)^40 chance of getting through.

Cheers

Magnus Westerlund

IETF Transport Area Director & TSVWG Chair
----------------------------------------------------------------------
Multimedia Technologies, Ericsson Research EAB/TVM/M
----------------------------------------------------------------------
Ericsson AB                | Phone +46 8 4048287
Torshamsgatan 23           | Fax   +46 8 7575550
S-164 80 Stockholm, Sweden | mailto: magnus.westerlund@ericsson.com
----------------------------------------------------------------------