From owner-freebsd-current@FreeBSD.ORG Sat Sep 13 13:26:59 2003 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0FE5916A4BF for ; Sat, 13 Sep 2003 13:26:59 -0700 (PDT) Received: from thuis.piwebs.com (217-19-20-186.dsl.cambrium.nl [217.19.20.186]) by mx1.FreeBSD.org (Postfix) with SMTP id 8EC4B43FBD for ; Sat, 13 Sep 2003 13:26:57 -0700 (PDT) (envelope-from avleeuwen@piwebs.com) Received: (qmail 47382 invoked by uid 85); 13 Sep 2003 20:28:13 -0000 Received: from avleeuwen@piwebs.com by thuis.piwebs.com by uid 82 with qmail-scanner-1.20rc1 (uvscan: v4.2.40/v4288. Clear:RC:1:. Processed in 0.132238 secs); 13 Sep 2003 20:28:13 -0000 Received: from unknown (HELO 192.168.0.109) (192.168.0.109) by 0 with SMTP; 13 Sep 2003 20:28:12 -0000 From: Arjan van Leeuwen To: Sebastian Ssmoller Date: Sat, 13 Sep 2003 22:26:54 +0200 User-Agent: KMail/1.5.3 References: <200309131913.33506.avleeuwen@piwebs.com> <4162.1063480375@www18.gmx.net> In-Reply-To: <4162.1063480375@www18.gmx.net> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200309132226.54360.avleeuwen@piwebs.com> cc: current@freebsd.org Subject: Re: Bad performance X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 13 Sep 2003 20:26:59 -0000 > > > > > > as mentioned: really bad performace occurs when lauching mozilla, gaim, > > > gnome2, etc. ... when mozilla is running the perfomance seems to be ok > > > ... possibly a bit too slow but i do not know how to proof this (?) > > > > > > i suppose a udma/disk/controller problem. i found out that the > > > southbridge i use (VIA 82C686B) has some bugs. but in fbsd 5.0 release > > > notes i found a bugfix for that so i am not sure about it ... > > > > I have the same southbridge and a very fast system. That can't be the > > problem > > do u have enabled/disabled anything special ? (kernel, io, net, ...) I have all debugging options in the kernel disabled, and I have a non-debugging malloc (ln -s aj /etc/malloc.conf), but this is disabled by default on 5.1-RELEASE I think (are you running -RELEASE or -CURRENT?). > > > - and it really does sound like a network problem. It looks like it's > > looking > > for a host that it can't find. Possibly your own hostname. > > ok. but what i do not understand is that when i do some tests manually > everything seems to be ok (e.g. ping). any ideas what i may test to figure > out > whether it is network problem or not ? > > btw. i did a simple io test: (ufs2 softupdates) > time dd if=/dev/zero of=./out bs=1024k count=256 > > 256+0 records in > 256+0 records out > 268435456 bytes transferred in 20.389193 secs (13165575 bytes/sec) > > real 0m20.401s > user 0m0.000s > sys 0m6.732s For the record, this is my output: 256+0 records in 256+0 records out 268435456 bytes transferred in 7.681891 secs (34943929 bytes/sec) dd if=/dev/zero of=./out bs=1024k count=256 0.00s user 3.66s system 47% cpu 7.772 total OK, so it's a little more than twice as fast, but I probably have a faster machine and a faster hard drive. This seems to me like it has nothing to do with the ultra-long startup times in GNOME. I've seen them before, and they were all related to network issues. Maybe you can try writing a small C program that does a gethostbyname on the output of gethostname and a gethostbyaddr on the output of gethostbyname and see if it works. > when i did this under linux it finished in no time. i do not know whether > that has to say anything but i found that rather interesting. They probably have different ways to write zeroes :). Arjan