Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262334AbVBKUnN (ORCPT ); Fri, 11 Feb 2005 15:43:13 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262332AbVBKUnN (ORCPT ); Fri, 11 Feb 2005 15:43:13 -0500 Received: from [83.102.214.158] ([83.102.214.158]:3473 "EHLO gw.home.net") by vger.kernel.org with ESMTP id S262334AbVBKUmu (ORCPT ); Fri, 11 Feb 2005 15:42:50 -0500 X-Comment-To: Mingming Cao To: Mingming Cao Cc: Andreas Dilger , Andrew Morton , tytso@mit.edu, pbadari@us.ibm.com, suparna@in.ibm.com, gerrit@us.ibm.com, tappro@clusterfs.com, linux-kernel , ext2-devel@lists.sourceforge.net, alex@clusterfs.com Subject: Re: Latest ext3 patches (extents, mballoc, delayed allocation) References: <1106354192.3634.19.camel@dyn318043bld.beaverton.ibm.com> <4207BBEA.7090705@us.ibm.com> From: Alex Tomas Organization: HOME User-Agent: Gnus/5.110002 (No Gnus v0.2) Emacs/21.3 (gnu/linux) Date: Fri, 11 Feb 2005 23:40:21 +0300 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 11337 Lines: 210 Good day all, I've updated the patchset against 2.6.10. A bunch of bugs have been fixed and mballoc now behaves smarter a bit. Extents and mballoc patches collects some stats they print upon umount. NOTE: they must not be used to store important data. A lot of things are to be done. Please review. Any comments and suggestions are very welcome. The patches are too big to send in a message (previous time they got rejected). I put them in ftp://ftp.clusterfs.com/pub/people/alex/2.6.10 The following names are used: ext3rs - ext3 mounted with data=writeback,reservation options ext3i - ext3 mounted with extents,mballoc,delalloc options reiser - reiserfs v3 I did some benchmarking. Two systems were used: SMP - dual PIII-1GHz, 1GB, FC-connected to 2 disks raid0 UP - PIII-933MHz, 256MB, FC-connected to 2 disks raid0 The tests dd500 and dd1000 generate 500M and 1000M using dd on fresh fs. Real time varies run from run because of disks, but sys.time is stable. SMP UP TEST / FS real sys real sys dd500/ext2 - 9.14 2.33 19.03 2.50 dd500/ext3 - 12.23 4.06 18.99 4.19 dd500/ext3rs - 13.39 4.01 15.08 4.16 dd500/ext3i - 9.19 2.31 19.07 2.52 dd500/reiser - 7.95 2.87 21.23 3.09 dd500/xfs - 17.88 2.42 19.25 2.67 dd1000/ext2 - 37.47 4.69 44.59 5.02 dd1000/ext3 - 38.03 7.90 40.77 8.31 dd1000/ext3rs - 34.62 7.95 40.51 8.30 dd1000/ext3i - 33.73 4.65 38.74 4.93 dd1000/reiser - 29.29 5.79 44.88 6.08 dd1000/xfs - 37.11 5.03 39.98 5.27 The test untar unpacks linux-2.6.0.tar: SMP UP TEST / FS real sys real sys untar/ext2 - 9.21 2.47 27.31 2.57 untar/ext3 - 15.63 3.38 34.54 3.61 untar/ext3rs - 15.91 3.27 35.98 3.65 untar/ext3i - 8.33 2.70 16.75 2.84 untar/reiser - 13.38 5.47 25.88 5.96 untar/xfs - 44.62 5.64 51.92 4.88 The next test is dbench: SMP UP TEST / FS real sys MB/s real sys MB/s dbench1/ext2 - 5.87 1.43 63.7256 5.93 1.54 63.79 dbench1/ext3 - 2.46 1.75 139.794 3.60 1.85 131.372 dbench1/ext3rs - 2.46 1.73 140.378 3.55 1.85 133.071 dbench1/ext3i - 2.19 1.48 156.976 2.28 1.53 150.403 dbench1/reiser - 2.80 2.05 122.761 2.88 2.11 119.815 dbench1/xfs - 2.83 1.81 122.159 2.87 1.91 119.564 dbench4/ext2 - 4.99 7.13 274.216 9.96 6.22 137.316 dbench4/ext3 - 5.79 8.64 236.858 11.49 7.40 130.146 dbench4/ext3rs - 5.80 8.57 236.356 11.45 7.38 130.467 dbench4/ext3i - 5.16 7.16 265.621 9.25 6.31 147.872 dbench4/reiser - 7.11 10.85 192.491 11.85 8.59 115.392 dbench4/xfs - 6.60 8.88 207.599 11.98 7.69 114.177 dbench8/ext2 - 16.04 14.27 181.93 18.32 12.14 149.228 dbench8/ext3 - 18.87 17.04 165.214 23.77 14.88 119.995 dbench8/ext3rs - 11.52 17.21 237.647 23.35 15.04 123.088 dbench8/ext3i - 11.20 14.66 268.052 21.00 12.49 130.223 dbench8/reiser - 13.92 21.65 196.306 24.98 17.67 114.083 dbench8/xfs - 14.84 18.23 184.263 25.84 15.53 105.743 dbench16/ext2 - 20.39 28.79 267.979 47.37 25.13 115.582 dbench16/ext3 - 25.69 34.54 212.745 53.43 30.78 102.266 dbench16/ext3rs - 24.47 34.36 223.37 54.68 30.44 100.082 dbench16/ext3i - 22.94 29.40 238.215 44.89 25.73 121.962 dbench16/reiser - 28.56 43.86 191.407 56.48 36.41 96.9686 dbench16/xfs - 33.49 36.94 163.159 78.68 32.56 69.4764 dbench32/ext2 - 43.54 58.87 250.947 108.1 51.24 101.354 dbench32/ext3 - 61.83 70.66 176.707 139.84 63.77 78.8818 dbench32/ext3rs - 67.24 71.69 162.651 145.03 63.38 75.5155 dbench32/ext3i - 55.29 59.41 197.757 100.24 52.50 109.26 dbench32/reiser - 76.32 93.45 143.127 128.77 77.94 85.7179 dbench32/xfs - 119.4 88.09 91.4746 670.45 76.19 16.298 The followins crazy listing shows tiobench's results for SMP box: Sequential Reads File Blk Num Avg CPU Identifier Size Size Thr Rate (CPU%) Latency Eff ---------------------------- ------ ----- --- ------ ------ --------- ----- ext2 512 4096 1 40.95 12.80% 0.095 320 ext3 512 4096 1 45.03 13.99% 0.086 322 ext3rs 512 4096 1 50.53 15.18% 0.077 333 ext3i 512 4096 1 49.69 14.53% 0.078 342 reiser 512 4096 1 29.56 12.08% 0.132 245 xfs 512 4096 1 44.58 15.25% 0.087 292 ext2 512 4096 2 22.36 13.21% 0.287 169 ext3 512 4096 2 45.94 23.48% 0.108 196 ext3rs 512 4096 2 46.29 23.57% 0.119 196 ext3i 512 4096 2 46.96 23.83% 0.111 197 reiser 512 4096 2 23.05 14.89% 0.252 155 xfs 512 4096 2 55.98 30.65% 0.064 183 ext2 512 4096 4 45.39 55.10% 0.155 82 ext3 512 4096 4 59.16 48.85% 0.103 121 ext3rs 512 4096 4 54.35 46.83% 0.215 116 ext3i 512 4096 4 65.81 55.53% 0.119 119 reiser 512 4096 4 26.35 32.38% 0.369 81 xfs 512 4096 4 71.81 66.22% 0.113 108 Random Reads File Blk Num Avg CPU Identifier Size Size Thr Rate (CPU%) Latency Eff ---------------------------- ------ ----- --- ------ ------ --------- ----- ext2 512 4096 1 119.05 40.37% 0.031 295 ext3 512 4096 1 134.78 37.08% 0.028 363 ext3rs 512 4096 1 25.18 8.377% 0.154 301 ext3i 512 4096 1 144.45 48.98% 0.026 295 reiser 512 4096 1 33.53 13.51% 0.115 248 xfs 512 4096 1 83.14 27.66% 0.045 301 ext2 512 4096 2 34.28 19.08% -0.290 180 ext3 512 4096 2 239.35 168.4% -0.549 142 ext3rs 512 4096 2 257.29 167.9% -0.376 153 ext3i 512 4096 2 174.08 126.9% -1.885 137 reiser 512 4096 2 45.32 25.80% -0.477 176 xfs 512 4096 2 26.27 16.80% -0.762 156 ext2 512 4096 4 138.70 173.9% -1.166 80 ext3 512 4096 4 159.41 141.7% -1.310 112 ext3rs 512 4096 4 89.80 87.91% -0.091 102 ext3i 512 4096 4 135.86 122.5% -3.192 111 reiser 512 4096 4 109.81 153.8% -0.673 71 xfs 512 4096 4 32.64 39.26% -1.949 83 Sequential Writes File Blk Num Avg CPU Identifier Size Size Thr Rate (CPU%) Latency Eff ---------------------------- ------ ----- --- ------ ------ --------- ----- ext2 512 4096 1 55.15 37.13% 0.030 149 ext3 512 4096 1 35.24 43.82% 0.106 80 ext3rs 512 4096 1 34.24 42.86% 0.125 80 ext3i 512 4096 1 43.54 32.07% 0.028 136 reiser 512 4096 1 51.67 84.37% 0.063 61 xfs 512 4096 1 27.32 21.85% 0.057 125 ext2 512 4096 2 18.76 34.45% 0.111 54 ext3 512 4096 2 28.82 88.31% 0.236 33 ext3rs 512 4096 2 30.63 91.98% 0.218 33 ext3i 512 4096 2 56.24 112.1% 0.018 50 reiser 512 4096 2 18.31 85.35% 0.173 21 xfs 512 4096 2 34.29 67.42% 0.213 51 ext2 512 4096 4 15.92 54.67% 0.559 29 ext3 512 4096 4 41.36 243.4% 0.137 17 ext3rs 512 4096 4 45.44 275.5% 0.197 16 ext3i 512 4096 4 55.94 208.8% 0.020 27 reiser 512 4096 4 20.47 176.9% 0.491 12 xfs 512 4096 4 53.60 207.4% 0.056 26 Random Writes File Blk Num Avg CPU Identifier Size Size Thr Rate (CPU%) Latency Eff ---------------------------- ------ ----- --- ------ ------ --------- ----- ext2 512 4096 1 0.73 0.565% 0.016 129 ext3 512 4096 1 1.01 0.733% 0.018 137 ext3rs 512 4096 1 1.01 0.742% 0.018 136 ext3i 512 4096 1 0.84 0.611% 0.015 137 reiser 512 4096 1 0.74 0.531% 0.114 138 xfs 512 4096 1 1.40 1.048% 0.018 134 ext2 512 4096 2 1.21 2.248% 0.021 54 ext3 512 4096 2 1.29 2.549% 0.021 51 ext3rs 512 4096 2 1.20 2.448% 0.021 49 ext3i 512 4096 2 0.74 1.572% 0.018 47 reiser 512 4096 2 1.28 2.325% 0.021 55 xfs 512 4096 2 1.08 2.056% 0.020 52 ext2 512 4096 4 1.04 3.760% 0.028 28 ext3 512 4096 4 0.76 2.758% 0.029 28 ext3rs 512 4096 4 0.80 3.052% 0.027 26 ext3i 512 4096 4 0.76 2.879% 0.024 26 reiser 512 4096 4 1.13 3.854% 0.023 29 xfs 512 4096 4 0.79 2.432% 0.025 32 thanks, Alex - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/