From owner-freebsd-hackers  Sat Mar 23 22:40:59 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id WAA05958
          for hackers-outgoing; Sat, 23 Mar 1996 22:40:59 -0800 (PST)
Received: from asstdc.scgt.oz.au (root@asstdc.scgt.oz.au [202.14.234.65])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id WAA05951
          for <freebsd-hackers@FreeBSD.ORG>; Sat, 23 Mar 1996 22:40:56 -0800 (PST)
Received: (from imb@localhost)
	by asstdc.scgt.oz.au (8.6.12/BSD4.4)
	id RAA18197; Sun, 24 Mar 1996 17:39:36 +1100
From: michael butler <imb@scgt.oz.au>
Message-Id: <199603240639.RAA18197@asstdc.scgt.oz.au>
Subject: Re: Changing Ethernet frame size to 576 bytes?
To: rkw@dataplex.net (Richard Wackerbarth)
Date: Sun, 24 Mar 1996 17:39:36 +1100 (EST)
Cc: taob@io.org, freebsd-hackers@FreeBSD.ORG
In-Reply-To: <n1384500272.3953@Richard Wackerbarth> from "Richard Wackerbarth" at Mar 23, 96 11:03:34 pm
X-Mailer: ELM [version 2.4 PL24beta]
Content-Type: text
Sender: owner-hackers@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

Richard Wackerbarth writes:

> His suggestion is valid. If the MTU is larger than the minimum MTU along
> the path, each packet that you send out will be broken into sub-packets
> and reassembled on the far end.

AFAIK this can happen two ways .. MTU discovery or at some (arbitrary)
router in the transit path.

> If any subpacket gets lost, the whole packet is lost and must be
> retransmitted.

Yup .. similar to the NFS 8k datagram disease suffered by hosts with slow
ethernet interfaces :-(
 
> Since tcp knows how to recover lost packets, the file eventually gets
> through. However, it is more efficient if your packets are not fragmented
> in transit.

Around here, anyone who has a BRI or better connection to the 'net runs with
an MTU of 1500. There's no point in crippling every site on the net because
some, hopefully few, transit paths can't cope,

	michael