From owner-freebsd-performance@FreeBSD.ORG Tue Jul 24 11:18:41 2007 Return-Path: Delivered-To: freebsd-performance@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id BEDB616A418 for ; Tue, 24 Jul 2007 11:18:41 +0000 (UTC) (envelope-from nico-freebsd-performance@schottelius.org) Received: from denkbrett.schottelius.org (natgw.netstream.ch [62.65.128.28]) by mx1.freebsd.org (Postfix) with ESMTP id 505DB13C458 for ; Tue, 24 Jul 2007 11:18:41 +0000 (UTC) (envelope-from nico-freebsd-performance@schottelius.org) Received: by denkbrett.schottelius.org (Postfix, from userid 1000) id AB19018841B; Tue, 24 Jul 2007 13:07:21 +0200 (CEST) Date: Tue, 24 Jul 2007 13:07:21 +0200 From: Nico -telmich- Schottelius To: freebsd-performance@freebsd.org Message-ID: <20070724110721.GA12779@schottelius.org> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="NzB8fVQJ5HfG6fxh" Content-Disposition: inline User-Agent: echo $message | gpg -e $sender -s | netcat mailhost 25 X-Unix-Info: http://unix.schottelius.org/ X-Netzseite: http://nico.schottelius.org/ X-System-Info: denkbrett running Linux 2.6.22.1-rsbac-denkbrett on i686 Subject: Very slow system using dd / nfs X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 24 Jul 2007 11:18:41 -0000 --NzB8fVQJ5HfG6fxh Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Hello guys, I am testing FreeBSD 6.2 NFS support, using linux-kernel-nfs as server. Now I am starting two instances of testdisks.sh [0] with forks=3D500 and directory set to /nfs/teufeli2/ (/nfs =3D nfs volume, teufeli2 is the host name). After some minutes the system has about 1500 processes running (which is pretty ok), but is extremly slow: - typing characters is delayed for about 10-20 seconds - typing ctrl+a " in screen takes about 30 seconds to display the menu - changing virtual consoles is fast (via ctrl+f[n]) - login in is still possible, though sometimes it takes longer than 300 seconds, which leads to a timeout I am wondering why the system behaves that way and what would be the correct way to priorize keyboard input for instance? The problem looks somehow similar to the problems I've seen with gmirror and our scsi disks (summary at [1]). Is it possible that freebsd has some quirks when having heavy i/o load in general? With heavy i/o load I mean about 500-1000 processes trying to access the disk system. I am currently trying to do some performance tests on freebsd to find out a) whether running a mailserver on a NFS volume would be possible (for a mailserver cluster) b) what todo to increase reaction time from a freebsd, so it's still maintainable under heavy i/o load c) what todo to increase mail server performance under freebsd in general? d) is nfs reliable under freebsd? Sincerly Nico [0]: http://home.schottelius.org/~nico/unix/freebsd/testdisks.sh [1]: Message-ID: <20070629111626.GF7663@schottelius.org> on freebsd-geom@ --=20 Think about Free and Open Source Software (FOSS). http://nico.schottelius.org/documentations/foss/the-term-foss/ PGP: BFE4 C736 ABE5 406F 8F42 F7CF B8BE F92A 9885 188C --NzB8fVQJ5HfG6fxh Content-Type: application/pgp-signature; name="signature.asc" Content-Description: Digital signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFGpd1puL75KpiFGIwRAk9PAJ9prOig5+wKO333Fpdp7Q+FIvGZqwCgnU7f 7M4DxXkJSWt2ayf8P/Om208= =v06B -----END PGP SIGNATURE----- --NzB8fVQJ5HfG6fxh-- From owner-freebsd-performance@FreeBSD.ORG Thu Jul 26 08:36:04 2007 Return-Path: Delivered-To: freebsd-performance@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 0DE2E16A417 for ; Thu, 26 Jul 2007 08:36:04 +0000 (UTC) (envelope-from markir@paradise.net.nz) Received: from smtp5.clear.net.nz (smtp5.clear.net.nz [203.97.33.68]) by mx1.freebsd.org (Postfix) with ESMTP id CEB1A13C442 for ; Thu, 26 Jul 2007 08:36:03 +0000 (UTC) (envelope-from markir@paradise.net.nz) Received: from zmori.markir.net (121-72-79-187.dsl.telstraclear.net [121.72.79.187]) by smtp5.clear.net.nz (CLEAR Net Mail) with ESMTP id <0JLS00C431UK1M30@smtp5.clear.net.nz> for freebsd-performance@freebsd.org; Thu, 26 Jul 2007 20:20:45 +1200 (NZST) Date: Thu, 26 Jul 2007 20:20:42 +1200 From: Mark Kirkwood In-reply-to: <458B3651.8090601@paradise.net.nz> To: freebsd-performance@freebsd.org Message-id: <46A8595A.7080400@paradise.net.nz> MIME-version: 1.0 Content-type: text/plain; charset=ISO-8859-1; format=flowed Content-transfer-encoding: 7bit References: <458B3651.8090601@paradise.net.nz> User-Agent: Thunderbird 2.0.0.4 (X11/20070716) Subject: Re: Cached file read performance X-BeenThere: freebsd-performance@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Performance/tuning List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 26 Jul 2007 08:36:04 -0000 I thought it might be interesting see how 7-CURRENT did with respect to cached file reading that we tested a while ago. Briefly recall that I'm using a dual PIII 1.26Ghz with 2G dual channel PC133, reading a 781MB (completely) cached file. Redoing the same test, with the previous 6.2-PRE results included for context, shows that we seem to be noticeably faster in 7 generally - which is nice. There is one regression - random reads for the 64K block size are slower than in 6.2-PRE. The results seem pretty stable... several runs yeild pretty much same ballpark figures (enough to say something like "10% faster sequential reads with 32K blocks") It would be interesting to compare with 6.2-STABLE as well... but I don't have one of the same machines running it. The results ----------------- FreeBSD 6.2 PRERELEASE (20061127): -------- $ ./readtest /data0/dump/file 8192 0 random reads: 100000 of: 8192 bytes elapsed: 4.4477s io rate: 184186327 bytes/s $ ./readtest /data0/dump/file 8192 1 sequential reads: 100000 of: 8192 bytes elapsed: 1.9797s io rate: 413804878 bytes/s $ ./readtest /data0/dump/file 32768 0 random reads: 25000 of: 32768 bytes elapsed: 2.0076s io rate: 408040469 bytes/s $ ./readtest /data0/dump/file 32768 1 sequential reads: 25000 of: 32768 bytes elapsed: 1.7068s io rate: 479965034 bytes/s $ ./readtest /data0/dump/file 65536 0 random reads: 12500 of: 65536 bytes elapsed: 1.7856s io rate: 458778279 bytes/s $ ./readtest /data0/dump/file 65536 1 sequential reads: 12500 of: 65536 bytes elapsed: 1.6611s io rate: 493158866 bytes/s FreeBSD 7.0-CURRENT (20070722): ------- $ ./readtest /data0/dump/file 8192 0 random reads: 100000 of: 8192 bytes elapsed: 3.8357s io rate: 213574160 bytes/s $ ./readtest /data0/dump/file 8192 1 sequential reads: 100000 of: 8192 bytes elapsed: 1.8215s io rate: 449743176 bytes/s $ ./readtest /data0/dump/file 32768 0 random reads: 25000 of: 32768 bytes elapsed: 1.8818s io rate: 435326721 bytes/s $ ./readtest /data0/dump/file 32768 1 sequential reads: 25000 of: 32768 bytes elapsed: 1.5504s io rate: 528388293 bytes/s $ ./readtest /data0/dump/file 65336 0 random reads: 12538 of: 65336 bytes elapsed: 1.9625s io rate: 417417120 bytes/s $ ./readtest /data0/dump/file 65336 1 sequential reads: 12538 of: 65336 bytes elapsed: 1.5365s io rate: 533140236 bytes/s regards Mark