From owner-freebsd-scsi@FreeBSD.ORG Thu Nov 6 05:20:28 2003 Return-Path: Delivered-To: freebsd-scsi@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id F025516A4CE for ; Thu, 6 Nov 2003 05:20:28 -0800 (PST) Received: from main.gmane.org (main.gmane.org [80.91.224.249]) by mx1.FreeBSD.org (Postfix) with ESMTP id 4F03943FCB for ; Thu, 6 Nov 2003 05:20:21 -0800 (PST) (envelope-from freebsd-scsi@m.gmane.org) Received: from root by main.gmane.org with local (Exim 3.35 #1 (Debian)) id 1AHk3f-0005kD-00 for ; Thu, 06 Nov 2003 14:20:19 +0100 X-Injected-Via-Gmane: http://gmane.org/ To: freebsd-scsi@freebsd.org Received: from sea.gmane.org ([80.91.224.252]) by main.gmane.org with esmtp (Exim 3.35 #1 (Debian)) id 1AHjsj-0005cE-00 for ; Thu, 06 Nov 2003 14:09:01 +0100 Received: from news by sea.gmane.org with local (Exim 3.35 #1 (Debian)) id 1AHjsj-0004C0-00 for ; Thu, 06 Nov 2003 14:09:01 +0100 From: Jesse Guardiani Date: Thu, 06 Nov 2003 08:08:59 -0500 Organization: WingNET Lines: 130 Message-ID: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7Bit X-Complaints-To: usenet@sea.gmane.org User-Agent: KNode/0.7.2 X-Mail-Copies-To: never Sender: news Subject: performance problem with amr driver and Perc 2/SC X-BeenThere: freebsd-scsi@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list Reply-To: jesse@wingnet.net List-Id: SCSI subsystem List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 06 Nov 2003 13:20:29 -0000 Howdy list, I have a 500Mhz Dell PowerEdge 4300 with: [7:22]jesse@chortos:[~]# uname -a FreeBSD chortos.wingnet.net 4.8-RELEASE FreeBSD 4.8-RELEASE #0: Fri Apr 11 12:59:08 EDT 2003 jesse@chortos.wingnet.net:/usr/src/sys/compile/CHORTOS i386 [7:42]jesse@chortos:[~]# grep amr /var/run/dmesg.boot amr0: mem 0xf4c00000-0xf4ffffff irq 10 at device 13.1 on pci2 amr0: Firmware 3.00, BIOS 1.36, 16MB RAM amrd0: on amr0 amrd0: 51834MB (106156032 sectors) RAID 5 (optimal) And I think I'm having performance problems. I am running a RAID 5 array with 5 18 Gb SCSI disks. The disks have varying speeds and capabilities: ------------------------------------------------------- Slot SCSI ID Model # or Serial Brand/Model Specs ---- ------- ----------------- ----------- ----- 0 (orig) 0 DNES-318350 E182115 S IBM Ultra 3 SCSI Wide (80 MB/Sec) P/N: 25L2142 18.2 GB 7,200 RPM 1 (orig) 1 DNES-318350 E182115 S IBM Ultra 3 SCSI Wide (80 MB/Sec) P/N: 25L2142 18.2 GB 7,200 RPM 2 2 ST318203LC Seagate/Cheetah Ultra 2 SCSI Wide (80 MB/Sec) 18.21 GB (5.4ms seek time) 10,000 RPM 3 3 Dell P/N: Quantum/Atlas 10K Ultra 3 SCSI (160 MB/Sec) JP-082YPV-12541-03R-00MB 18.2 GB TN18J462 10,000 RPM 4 4 --BLANK-- 5 (hotsp) 5 PN34L7404 IBM Ultra 3 SCSI (160 MB/Sec) ECF24486 18 GB 7,200 RPM 8 8 --BLANK-- 9 9 --BLANK-- ------------------------------------------------------- The problem is that this box is averaging the following, according to tests with dd: write read 2.05 MByte/Sec 12.59 MByte/Sec Here are the exact tests I run: write test: sh -c 'time -h dd if=/dev/zero bs=1024k of=tstfile count=1024' read test: sh -c 'time -h dd if=tstfile bs=1024k of=/dev/null' Each test is run seperately. I monitor CPU usage via top in a seperate terminal and monitor `iostat -w 1` in yet another terminal. Here is the result of the write test: 1024+0 records in 1024+0 records out 1073741824 bytes transferred in 498.681822 secs (2153160 bytes/sec) 8m19.03s real 0.01s user 17.41s sys CPU varied during this test, but I'd guess that I averaged 50% idle or more. This is typically what the dd process looked like in top: 45270 root -18 0 2268K 1128K wdrain 0:03 3.16% 3.03% dd Here is some of the iostat output I get during the write test: ------------------------------------------------------- tty amrd0 cd0 sa0 cpu tin tout KB/t tps MB/s KB/t tps MB/s KB/t tps MB/s us ni sy in id 0 227 60.05 41 2.38 0.00 0 0.00 0.00 0 0.00 56 0 42 2 0 3 6142 53.31 49 2.53 0.00 0 0.00 0.00 0 0.00 26 0 16 2 57 0 377 71.95 38 2.64 0.00 0 0.00 0.00 0 0.00 1 0 7 2 91 0 369 88.18 34 2.90 0.00 0 0.00 0.00 0 0.00 1 0 5 0 95 0 341 57.42 38 2.11 0.00 0 0.00 0.00 0 0.00 1 0 5 1 93 0 2285 58.48 42 2.37 0.00 0 0.00 0.00 0 0.00 1 0 10 1 88 0 2420 79.33 36 2.76 0.00 0 0.00 0.00 0 0.00 0 0 5 0 95 0 2393 88.24 34 2.89 0.00 0 0.00 0.00 0 0.00 2 0 7 0 92 0 337 58.29 42 2.37 0.00 0 0.00 0.00 0 0.00 4 0 10 0 86 0 354 70.91 35 2.40 0.00 0 0.00 0.00 0 0.00 3 0 6 2 89 0 419 50.91 44 2.17 0.00 0 0.00 0.00 0 0.00 3 0 6 0 91 0 76 39.90 60 2.35 0.00 0 0.00 0.00 0 0.00 12 0 10 0 78 0 1181 60.49 44 2.61 0.00 0 0.00 0.00 0 0.00 19 0 42 2 36 0 76 96.00 31 2.88 0.00 0 0.00 0.00 0 0.00 40 0 49 0 12 0 2629 68.82 39 2.60 0.00 0 0.00 0.00 0 0.00 35 0 13 2 51 0 416 77.73 30 2.25 0.00 0 0.00 0.00 0 0.00 1 0 9 0 91 0 2509 88.77 26 2.23 0.00 0 0.00 0.00 0 0.00 1 0 5 1 94 0 299 35.26 53 1.84 0.00 0 0.00 0.00 0 0.00 0 0 5 1 94 0 2417 27.12 72 1.91 0.00 0 0.00 0.00 0 0.00 4 0 3 0 93 0 2463 22.20 81 1.76 0.00 0 0.00 0.00 0 0.00 11 0 14 0 75 ------------------------------------------------------- Any ideas on why I can't throughput more than 3 MB/s? Does it have anything to do with the fact that my disks have different RPM speeds? My 5400 RPM ATA laptop hard drive (~13 MB/Sec Read and Write) gets better throughput than this RAID 5 array. Something is very wrong. -- Jesse Guardiani, Systems Administrator WingNET Internet Services, P.O. Box 2605 // Cleveland, TN 37320-2605 423-559-LINK (v) 423-559-5145 (f) http://www.wingnet.net