From owner-freebsd-hackers  Thu Mar  7  0:53:47 2002
Delivered-To: freebsd-hackers@freebsd.org
Received: from gwdu60.gwdg.de (gwdu60.gwdg.de [134.76.98.60])
	by hub.freebsd.org (Postfix) with ESMTP id F1DB737B402
	for <freebsd-hackers@freebsd.org>; Thu,  7 Mar 2002 00:53:43 -0800 (PST)
Received: from localhost (kheuer@localhost)
	by gwdu60.gwdg.de (8.11.6/8.11.6) with ESMTP id g278rf719675;
	Thu, 7 Mar 2002 09:53:41 +0100 (CET)
	(envelope-from kheuer@gwdg.de)
X-Authentication-Warning: gwdu60.gwdg.de: kheuer owned process doing -bs
Date: Thu, 7 Mar 2002 09:53:40 +0100 (CET)
From: Konrad Heuer <kheuer@gwdg.de>
To: Dimitar Peikov <mitko@rila.bg>
Cc: freebsd-hackers@freebsd.org
Subject: Re: Swapping performance
In-Reply-To: <20020307104518.0f73740b.mitko@rila.bg>
Message-ID: <20020307094856.K17236-100000@gwdu60.gwdg.de>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=X-UNKNOWN
Content-Transfer-Encoding: QUOTED-PRINTABLE
Sender: owner-freebsd-hackers@FreeBSD.ORG
Precedence: bulk
List-ID: <freebsd-hackers.FreeBSD.ORG>
List-Archive: <http://docs.freebsd.org/mail/> (Web Archive)
List-Help: <mailto:majordomo@FreeBSD.ORG?subject=help> (List Instructions)
List-Subscribe: <mailto:majordomo@FreeBSD.ORG?subject=subscribe%20freebsd-hackers>
List-Unsubscribe: <mailto:majordomo@FreeBSD.ORG?subject=unsubscribe%20freebsd-hackers>
X-Loop: FreeBSD.ORG


On Thu, 7 Mar 2002, Dimitar Peikov wrote:

> I start some performance tests on -stable and on SuSE 7.1 / 2.4.17. I
> don't comment about 'bzero' performance, but when RAM is over, Linux
> is much faster. I have no idea what is the algorithm of swapping but it s=
eems that the granularity of swapping pieces is the key or the importance o=
f swapping memory blocks of certain task. Ooo I forgot to say that the both=
 machines have the same hardware, IBM 300PL, 256 RAM and no other tasks run=
ning. I had to run these tests to choose the fastest platform for building =
our software indexes, which requires a lot of math and memory operations.
>
> --- with bzero ---
> Linux$ time ./malloc_test
> *#
> real    0m37.640s
> user    0m1.370s
> sys     0m2.950s
> Linux$
>
> FreeBSD$ time ./malloc_test
> *#
> real    0m46.640s
> user    0m2.280s
> sys     0m2.550s
> FreeBSD$
>
> --- without bzero ---
> Linux$ time ./malloc_test
> *#
> real    0m6.371s
> user    0m0.450s
> sys     0m1.510s
> Linux$
>
> FreeBSD$ time ./malloc_test
> *#
> real    0m11.571s
> user    0m1.150s
> sys     0m1.830s
> FreeBSD$

Just to make sure: What about disk layout and paging space location? Both
systems will behave best when paging space location is near to the
"beginning" of the disks.

My measurements in this area are some years old; at that time FreeBSD did
a much better job when klow on free memory.

Best regards
Konrad

Konrad Heuer                                    Personal Bookmarks:
Gesellschaft f=FCr wissenschaftliche
   Datenverarbeitung mbH G=D6ttingen              http://www.freebsd.org
Am Fa=DFberg, D-37077 G=D6ttingen                   http://www.daemonnews.o=
rg
Deutschland (Germany)

kheuer@gwdg.de


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message