[3787] in linux-net channel archive

home help back first fref pref prev next nref lref last post

Re: More data on wizard MTU/Fragmentation problem

daemon@ATHENA.MIT.EDU (Alan Cox)
Sat Jul 20 16:58:34 1996

From: iialan@iifeak.swan.ac.uk (Alan Cox)
To: gpa@piranha.ianet.net (Garry Adkins)
Date: 	Fri, 19 Jul 1996 10:03:57 +0100 (BST)
Cc: linux-net@vger.rutgers.edu
In-Reply-To: <199607171239.IAA20017@piranha.ianet.net> from "Garry Adkins" at Jul 17, 96 08:39:51 am

> This is similar to our original manefestion here.  Our News server
> (linux 1.2.8) suddenly became unreachable.  I was able to kludge a 
> fix for that by changing the eth0 interface to use the smallest
> MTU that any customer had (576).  Now packets don't need to be
> fragmented.  I originally planned to simply change the MTU on
> my cisco router to the same, but it doesn't support changing
> the mtu on the ethernet port.  So I'm still kind of stuck.

The linux PPP code has a fragment bug (the bug is in ip_fragment) but its
for all but amateur radio users explicitly a PPP bug. That is fixed in the
patches sent on to Linus for 2.0.8 [and kudos to the person who found and
fixed it]

The other thing you will see repeatedly if you have SLIP users with less than
10 brain cells or poor setup is MTU discovery problems. All those users who
had trouble with some near web sites only (normally those running Solaris 2.4/5)
over their typically SLIP dialup will find Linux also starts to annoy them. 
Thats because they have differing MTU settings each end of the SLIP link. Linux
like Solaris now uses MTU discovery, and that does require you set the thing
up right. Previously that sort of error would sometimes hide.

Alan



home help back first fref pref prev next nref lref last post