[4164] in linux-net channel archive
Re: TCP/buffering changes
daemon@ATHENA.MIT.EDU (Darren Crane)
Fri Aug 23 15:57:04 1996
Date: Fri, 23 Aug 1996 14:04:55 -0400 (EDT)
From: Darren Crane <dcrane@ivs.piedmont.net>
To: Alan Cox <alan@lxorguk.ukuu.org.uk>
cc: linux-kernel@vger.rutgers.edu, linux-net@vger.rutgers.edu,
torvalds@cs.helsinki.fi
In-Reply-To: <m0uth6Y-0005FcC@lightning.swansea.linux.org.uk>
Something between 2.0.10 and 2.0.11 slowed samba down.
I'm running samba with the following socket options...
read prediction = True
read size = 4096
socket options = IPTOS_LOWDELAY TCP_NODELAY SO_SNDBUF=4096
SO_RCVBUF=32768
Why a SO_SNDBUF of 4k? 90% of trafic is from our home grown dos app
which reads files in 4K chunks. Why? Cause Dos sucks. But the slowdown
I experience happens with regular dos copy commands as well, and it still
exists if I up the SO_SNDBUF to 32768 (or anything between the two).
Darren Crane
dcrane@ivs.piedmont.net
On Thu, 22 Aug 1996, Alan Cox wrote:
>Any chance of trying say 2.0.0 and 2.0.8 to try and narrow this down. It
>does look interesting and probably like a bug. Either that or we speeded it
>up too much for your clients ;). Somehow I suspect we have some kind of
>interaction of the former kind. Do you have the TCP_NODELAY option enabled
>in SAMBA ?
>
>Alan
>
