https://raid.wiki.kernel.org/index.php?title=Performance&feed=atom&action=historyPerformance - Revision history2024-03-28T18:48:40ZRevision history for this page on the wikiMediaWiki 1.19.24https://raid.wiki.kernel.org/index.php?title=Performance&diff=2542&oldid=prevKeld: /* Other benchmarks from 2007-2010 */2011-05-05T11:07:07Z<p><span dir="auto"><span class="autocomment">Other benchmarks from 2007-2010</span></span></p>
<table class='diff diff-contentalign-left'>
<col class='diff-marker' />
<col class='diff-content' />
<col class='diff-marker' />
<col class='diff-content' />
<tr valign='top'>
<td colspan='2' style="background-color: white; color:black;">← Older revision</td>
<td colspan='2' style="background-color: white; color:black;">Revision as of 11:07, 5 May 2011</td>
</tr><tr><td colspan="2" class="diff-lineno">Line 147:</td>
<td colspan="2" class="diff-lineno">Line 147:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>Keld's remarks: the raid0 read figure of 110 MB/s is not consistent with other benchmarks that report about cumulative performance for sequential reads for raid0. I would have expected a figure around 150 MB/s here.</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>Keld's remarks: the raid0 read figure of 110 MB/s is not consistent with other benchmarks that report about cumulative performance for sequential reads for raid0. I would have expected a figure around 150 MB/s here.</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>In July 2008, Jon Nelson made [http://blog.jamponi.net/2008/07/raid56-and-10-benchmarks-on-26255_10.html test of levels 5, 6, and 10, with differing chunk sizes from 64 kiB to 2 MiB] on 4 SATA drives. This was for sequential read and writes on the raw raid types. With a file system, differences in write <del class="diffchange diffchange-inline">perfornance </del>would probably smoothen out write differences due to the effect of the IO scheduler (elevator). Results include high performance of raid10,f2 - around 3.80, and high performance of raid5 and <del class="diffchange diffchange-inline">rsid6</del>. Especially with bigger chunk sizes 512 kiB - 2 MiB, raid5 obtained maximum 3.44 times the speed of a single drive, and raid6 got a factor 2.96. This is probably due to the even distribution of parity chunks, that means that reads are distributed evenly on all involved (4) drives.</div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>In July 2008, Jon Nelson made [http://blog.jamponi.net/2008/07/raid56-and-10-benchmarks-on-26255_10.html test of levels 5, 6, and 10, with differing chunk sizes from 64 kiB to 2 MiB] on 4 SATA drives. This was for sequential read and writes on the raw raid types. With a file system, differences in write <ins class="diffchange diffchange-inline">performance </ins>would probably smoothen out write differences due to the effect of the IO scheduler (elevator). Results include high performance of raid10,f2 - around 3.80, and high performance of raid5 and <ins class="diffchange diffchange-inline">raid6</ins>. Especially with bigger chunk sizes 512 kiB - 2 MiB, raid5 obtained maximum 3.44 times the speed of a single drive, and raid6 got a factor 2.96. This is probably due to the even distribution of parity chunks, that means that reads are distributed evenly on all involved (4) drives.</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>In July 2008, Ben Martin made a test comparing HW and SW raid with 6 disks. The HW raid was a quite expensive USD 800 Adaptec SAS-31205 PCI Express 12-SATA-port PCI-E x8 hardware RAID card. Compared raid types included [http://www.linux.com/feature/140734|RAID 5, 6 and 10,n2]. Some conclusions: The difference is not big between the expensive HW raid controller and Linux SW RAID. For raid5 Linux was 30 % faster (440 MB/s vs 340 MB/s) for reads. For writes Adaptec <del class="diffchange diffchange-inline">swas </del>about 25 % faster (220 MB/s vs 175 MB/s). Keld's remarks: The Adaptec controller actually slowed down disk reading, the single disk read on the motherboard was 90 MB/s while via the Adaptec controller it was only 70 MB/s. Writing was faster, tho. raid10,f2 and raid10,o2 was not included in the test. Ben reported the Adaptec controller to give around 310 MB/s read performance for raid1, while a raid10,f2 would probably have given around 400 MB/s via the Adaptec controller, and around 600 MB/s with the 6 disks on the motherboard SATA controller and a reasonable extra controller, given other benchmarks as noted in this section, giving raid10,f2 a 95 % utilization of the cumulated IO bandwidth. For raid5 the read/write difference could possibly be explained by which chunk size to use, in Linux raid5 reading improves with bigger chunk sizes, while writing degrades.</div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>In July 2008, Ben Martin made a test comparing HW and SW raid with 6 disks. The HW raid was a quite expensive USD 800 Adaptec SAS-31205 PCI Express 12-SATA-port PCI-E x8 hardware RAID card. Compared raid types included [http://www.linux.com/feature/140734|RAID 5, 6 and 10,n2]. Some conclusions: The difference is not big between the expensive HW raid controller and Linux SW RAID. For raid5 Linux was 30 % faster (440 MB/s vs 340 MB/s) for reads. For writes Adaptec <ins class="diffchange diffchange-inline">was </ins>about 25 % faster (220 MB/s vs 175 MB/s). Keld's remarks: The Adaptec controller actually slowed down disk reading, the single disk read on the motherboard was 90 MB/s while via the Adaptec controller it was only 70 MB/s. Writing was faster, tho. raid10,f2 and raid10,o2 was not included in the test. Ben reported the Adaptec controller to give around 310 MB/s read performance for raid1, while a raid10,f2 would probably have given around 400 MB/s via the Adaptec controller, and around 600 MB/s with the 6 disks on the motherboard SATA controller and a reasonable extra controller, given other benchmarks as noted in this section, giving raid10,f2 a 95 % utilization of the cumulated IO bandwidth. For raid5 the read/write difference could possibly be explained by which chunk size to use, in Linux raid5 reading improves with bigger chunk sizes, while writing degrades.</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>In 2009 [http://alephnull.com/benchmarks/sata2009/chunksize.html A Comparison of Chunk Size for Software RAID-5] was done by Rik Faith with chunk sizes of 4 KiB to 64 MiB. It was found that chunk sizes of 128 KiB gave the best overall performance. The test was done on a Supermicro AOC-SAT2-MV8 controller with 8 SATA II ports, and connected to a 32-bit PCI slot, which could explain the 130 MB/s max found.  </div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>In 2009 [http://alephnull.com/benchmarks/sata2009/chunksize.html A Comparison of Chunk Size for Software RAID-5] was done by Rik Faith with chunk sizes of 4 KiB to 64 MiB. It was found that chunk sizes of 128 KiB gave the best overall performance. The test was done on a Supermicro AOC-SAT2-MV8 controller with 8 SATA II ports, and connected to a 32-bit PCI slot, which could explain the 130 MB/s max found.  </div></td></tr>
</table>Keldhttps://raid.wiki.kernel.org/index.php?title=Performance&diff=2532&oldid=prevKeld: /* New benchmarks from 2011 */2011-05-05T10:50:26Z<p><span dir="auto"><span class="autocomment">New benchmarks from 2011</span></span></p>
<table class='diff diff-contentalign-left'>
<col class='diff-marker' />
<col class='diff-content' />
<col class='diff-marker' />
<col class='diff-content' />
<tr valign='top'>
<td colspan='2' style="background-color: white; color:black;">← Older revision</td>
<td colspan='2' style="background-color: white; color:black;">Revision as of 10:50, 5 May 2011</td>
</tr><tr><td colspan="2" class="diff-lineno">Line 57:</td>
<td colspan="2" class="diff-lineno">Line 57:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>=New benchmarks from 2011=</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>=New benchmarks from 2011=</div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">Remark from keld: The tests reported by Mathias B below is actually carried out in an environment with almost 100 % CPU utilization, so I am not sure how enlightening the numbers are. </ins></div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>Mathias B posted some benchmarks to the mailing list with this setup:</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>Mathias B posted some benchmarks to the mailing list with this setup:</div></td></tr>
</table>Keldhttps://raid.wiki.kernel.org/index.php?title=Performance&diff=2482&oldid=prevViridian Development Corporation: add navigation2011-04-04T11:30:08Z<p>add navigation</p>
<table class='diff diff-contentalign-left'>
<col class='diff-marker' />
<col class='diff-content' />
<col class='diff-marker' />
<col class='diff-content' />
<tr valign='top'>
<td colspan='2' style="background-color: white; color:black;">← Older revision</td>
<td colspan='2' style="background-color: white; color:black;">Revision as of 11:30, 4 April 2011</td>
</tr><tr><td colspan="2" class="diff-lineno">Line 1:</td>
<td colspan="2" class="diff-lineno">Line 1:</td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">{| style="border:1px solid #aaaaaa; background-color:#f9f9f9;width:100%; font-family: Verdana, sans-serif;"</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">|- padding:5px;padding-top:0.5em;font-size: 95%; </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">| Back to [[Growing]] <span style="float:right; padding-left:5px;">Forward to [[Related tools]]</span></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">|}</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>=Performance of raids with 2 disks=</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>=Performance of raids with 2 disks=</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td colspan="2" class="diff-lineno">Line 404:</td>
<td colspan="2" class="diff-lineno">Line 411:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>at RAID, but it can test more than 2TB of storage on 32-bit machines,</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>at RAID, but it can test more than 2TB of storage on 32-bit machines,</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>and tests for file creat, stat, unlink operations.</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>and tests for file creat, stat, unlink operations.</div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">{| style="border:1px solid #aaaaaa; background-color:#f9f9f9;width:100%; font-family: Verdana, sans-serif;"</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">|- padding:5px;padding-top:0.5em;font-size: 95%; </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">| Back to [[Growing]] <span style="float:right; padding-left:5px;">Forward to [[Related tools]]</span></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">|}</ins></div></td></tr>
</table>Viridian Development Corporationhttps://raid.wiki.kernel.org/index.php?title=Performance&diff=2242&oldid=prevMathiasb at 01:57, 19 February 20112011-02-19T01:57:00Z<p></p>
<table class='diff diff-contentalign-left'>
<col class='diff-marker' />
<col class='diff-content' />
<col class='diff-marker' />
<col class='diff-content' />
<tr valign='top'>
<td colspan='2' style="background-color: white; color:black;">← Older revision</td>
<td colspan='2' style="background-color: white; color:black;">Revision as of 01:57, 19 February 2011</td>
</tr><tr><td colspan="2" class="diff-lineno">Line 95:</td>
<td colspan="2" class="diff-lineno">Line 95:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>The 6 HDDs are in a RAID5 array, with LVM on top and then ext4 on top of that. This is a very low end system so many results are bottlenecked by the CPU. NCQ enabled on all drives. Here are some bonnie++ results:</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>The 6 HDDs are in a RAID5 array, with LVM on top and then ext4 on top of that. This is a very low end system so many results are bottlenecked by the CPU. NCQ enabled on all drives. Here are some bonnie++ results:</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>     ------Sequential Output------ --Sequential Input-     --Random- ------Sequential Create------ --------Random Create-------- <del class="diffchange diffchange-inline"> </del></div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>     ------Sequential Output------ --Sequential Input-     --Random- ------Sequential Create------ --------Random Create--------</div></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>     -Per Chr- --Block-- -Rewrite- -Per Chr- --Block--  --Seeks-- -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete-- <del class="diffchange diffchange-inline"> </del></div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>     -Per Chr- --Block-- -Rewrite- -Per Chr- --Block--  --Seeks-- -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--</div></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>     Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP <del class="diffchange diffchange-inline"> </del></div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>     Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP</div></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>     ion 7G 16086 90 161172 78 69515 34 17887 97 258229  43  424 2 25336 98 +++++ +++ 31240 99 26047 99 +++++ +++ 31072 99 <del class="diffchange diffchange-inline"> </del></div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>     ion 7G 16086 90 161172 78 69515 34 17887 97 258229  43  424 2 25336 98 +++++ +++ 31240 99 26047 99 +++++ +++ 31072 99</div></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>     ion 7G 17434 98 184348 89 147077 73 18282 99 382536  51  465.3 2 25401 99 +++++ +++ 28185 90 26125 99 +++++ +++ 31352 100 <del class="diffchange diffchange-inline"> </del></div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>     ion 7G 17434 98 184348 89 147077 73 18282 99 382536  51  465.3 2 25401 99 +++++ +++ 28185 90 26125 99 +++++ +++ 31352 100</div></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>     ion 7G 17303 98 186702 90 142494 70 18310 99 356491  49  467.7 2 25447 99 +++++ +++ 28446 90 25677 99 +++++ +++ 31305 99 <del class="diffchange diffchange-inline"> </del></div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>     ion 7G 17303 98 186702 90 142494 70 18310 99 356491  49  467.7 2 25447 99 +++++ +++ 28446 90 25677 99 +++++ +++ 31305 99</div></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>     ion 7G 17322 98 171309 89 146962 74 18348 99 365177  51  456.9 2 20419 80 +++++ +++ 31455 100 25966 99 +++++ +++ 31096 99 <del class="diffchange diffchange-inline"> </del></div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>     ion 7G 17322 98 171309 89 146962 74 18348 99 365177  51  456.9 2 20419 80 +++++ +++ 31455 100 25966 99 +++++ +++ 31096 99</div></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>     ion 7G 17359 98 184704 91 123822 57 18282 99 375295  49  463.2 2 24908 98 +++++ +++ 31465 99 25969 99 +++++ +++ 31285 99 <del class="diffchange diffchange-inline"> </del></div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>     ion 7G 17359 98 184704 91 123822 57 18282 99 375295  49  463.2 2 24908 98 +++++ +++ 31465 99 25969 99 +++++ +++ 31285 99</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>     ion 7G 17310 98 182821 90 124661 58 18347 99 385963  52  459.5 2 24710 98 +++++ +++ 31840 99 26162 98 +++++ +++ 31502 99</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>     ion 7G 17310 98 182821 90 124661 58 18347 99 385963  52  459.5 2 24710 98 +++++ +++ 31840 99 26162 98 +++++ +++ 31502 99</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
</table>Mathiasbhttps://raid.wiki.kernel.org/index.php?title=Performance&diff=2232&oldid=prevMathiasb: Added benchmark results from Mathias B2011-02-19T01:53:56Z<p>Added benchmark results from Mathias B</p>
<table class='diff diff-contentalign-left'>
<col class='diff-marker' />
<col class='diff-content' />
<col class='diff-marker' />
<col class='diff-content' />
<tr valign='top'>
<td colspan='2' style="background-color: white; color:black;">← Older revision</td>
<td colspan='2' style="background-color: white; color:black;">Revision as of 01:53, 19 February 2011</td>
</tr><tr><td colspan="2" class="diff-lineno">Line 48:</td>
<td colspan="2" class="diff-lineno">Line 48:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>My kernel did not allow me to test RAID10,o2 as this is only supported from kernel 2.6.18.</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>My kernel did not allow me to test RAID10,o2 as this is only supported from kernel 2.6.18.</div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">=New benchmarks from 2011=</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">Mathias B posted some benchmarks to the mailing list with this setup:</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Motherboard: Zotac ION Synergy DDR2 (Atom 330 overclocked to 2GHz, 667 FSB)</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    RAM: 4GB DDR2 PC5300</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    SATA controller: 05:00.0 SCSI storage controller: HighPoint Technologies, Inc. RocketRAID 230x 4 Port SATA-II            Controller (rev 02)</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    SATA controller: nVidia Corporation MCP79 AHCI Controller (rev b1)</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Hard drives:</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Model=WDC WD20EARS-00MVWB0, FwRev=51.0AB51</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Model=WDC WD20EARS-00MVWB0, FwRev=50.0AB50</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Model=WDC WD20EARS-00MVWB0, FwRev=50.0AB50</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Model=SAMSUNG HD204UI, FwRev=1AQ10003</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Model=WDC WD20EARS-00MVWB0, FwRev=51.0AB51</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Model=SAMSUNG HD204UI, FwRev=1AQ10003</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">3 of these are connected to the PCI-E (1.0) SATA HBA and thereby bottle necked. OS is Archlinux 64-bit, kernel 2.6.37.1 & mdadm 3.1.4.</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">md details:</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    /dev/md0:</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Version : 1.2</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Creation Time : Tue Oct 19 08:58:41 2010</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Raid Level : raid5</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Array Size : 9751756800 (9300.00 GiB 9985.80 GB)</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Used Dev Size : 1950351360 (1860.00 GiB 1997.16 GB)</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Raid Devices : 6</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    [...]</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Layout : left-symmetric</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Chunk Size : 64K</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">ext4 details:</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    RAID stride:              16</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    RAID stripe width:        80</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    rw,noatime,barrier=1,stripe=80,data=writeback</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">block details:</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    md0 block readahead 65536</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    lvm lv block readahead 16384</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    /sys/block/md0/md/stripe_cache_size 16384</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    /sys/block/md0/queue/read_ahead_kb 65536</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">The 6 HDDs are in a RAID5 array, with LVM on top and then ext4 on top of that. This is a very low end system so many results are bottlenecked by the CPU. NCQ enabled on all drives. Here are some bonnie++ results:</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    ------Sequential Output------ --Sequential Input-     --Random- ------Sequential Create------ --------Random Create-------- </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block--  --Seeks-- -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete-- </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    ion 7G 16086 90 161172 78 69515 34 17887 97 258229  43  424 2 25336 98 +++++ +++ 31240 99 26047 99 +++++ +++ 31072 99 </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    ion 7G 17434 98 184348 89 147077 73 18282 99 382536  51  465.3 2 25401 99 +++++ +++ 28185 90 26125 99 +++++ +++ 31352 100 </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    ion 7G 17303 98 186702 90 142494 70 18310 99 356491  49  467.7 2 25447 99 +++++ +++ 28446 90 25677 99 +++++ +++ 31305 99 </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    ion 7G 17322 98 171309 89 146962 74 18348 99 365177  51  456.9 2 20419 80 +++++ +++ 31455 100 25966 99 +++++ +++ 31096 99 </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    ion 7G 17359 98 184704 91 123822 57 18282 99 375295  49  463.2 2 24908 98 +++++ +++ 31465 99 25969 99 +++++ +++ 31285 99 </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">    ion 7G 17310 98 182821 90 124661 58 18347 99 385963  52  459.5 2 24710 98 +++++ +++ 31840 99 26162 98 +++++ +++ 31502 99</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;"></ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins style="color: red; font-weight: bold; text-decoration: none;">They vary a bit because I played with readahead and cache settings, ultimately I ended up with the settings posted above the results.</ins></div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>=Other benchmarks from 2007-2010=</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>=Other benchmarks from 2007-2010=</div></td></tr>
</table>Mathiasbhttps://raid.wiki.kernel.org/index.php?title=Performance&diff=2057&oldid=prevBobBagwill: /* Fresh benchmarking tools */ wikifying2010-08-11T16:32:58Z<p><span dir="auto"><span class="autocomment">Fresh benchmarking tools: </span> wikifying</span></p>
<table class='diff diff-contentalign-left'>
<col class='diff-marker' />
<col class='diff-content' />
<col class='diff-marker' />
<col class='diff-content' />
<tr valign='top'>
<td colspan='2' style="background-color: white; color:black;">← Older revision</td>
<td colspan='2' style="background-color: white; color:black;">Revision as of 16:32, 11 August 2010</td>
</tr><tr><td colspan="2" class="diff-lineno">Line 305:</td>
<td colspan="2" class="diff-lineno">Line 305:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>hdparm.  It won't do real benchmarking of the arrays.</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>hdparm.  It won't do real benchmarking of the arrays.</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>Instead of hdparm, take a look at the tools described here: IOzone and</div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>Instead of hdparm, take a look at the tools described here: <ins class="diffchange diffchange-inline">[http://www.iozone.org/ </ins>IOzone<ins class="diffchange diffchange-inline">] </ins>and <ins class="diffchange diffchange-inline">[http://sourceforge.net/projects/bonnie/ </ins>Bonnie++<ins class="diffchange diffchange-inline">]</ins>.</div></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>Bonnie++.</div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div></div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>IOzone is a small, versatile and modern tool to use. It benchmarks</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>IOzone is a small, versatile and modern tool to use. It benchmarks</div></td></tr>
</table>BobBagwillhttps://raid.wiki.kernel.org/index.php?title=Performance&diff=2029&oldid=prevKeld: /* Some problem solving for benchmarking */2010-07-26T11:28:24Z<p><span dir="auto"><span class="autocomment">Some problem solving for benchmarking</span></span></p>
<table class='diff diff-contentalign-left'>
<col class='diff-marker' />
<col class='diff-content' />
<col class='diff-marker' />
<col class='diff-content' />
<tr valign='top'>
<td colspan='2' style="background-color: white; color:black;">← Older revision</td>
<td colspan='2' style="background-color: white; color:black;">Revision as of 11:28, 26 July 2010</td>
</tr><tr><td colspan="2" class="diff-lineno">Line 98:</td>
<td colspan="2" class="diff-lineno">Line 98:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>  # Set read-ahead.</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>  # Set read-ahead.</div></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>  echo "Setting read-ahead to <del class="diffchange diffchange-inline">64 </del>MiB for /dev/md3"</div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>  echo "Setting read-ahead to <ins class="diffchange diffchange-inline">32 </ins>MiB for /dev/md3"</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>  blockdev --setra 65536 /dev/md3</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>  blockdev --setra 65536 /dev/md3</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
</table>Keldhttps://raid.wiki.kernel.org/index.php?title=Performance&diff=2028&oldid=prevKeld: /* Other benchmarks from 2007-2010 */2010-07-19T18:51:18Z<p><span dir="auto"><span class="autocomment">Other benchmarks from 2007-2010</span></span></p>
<table class='diff diff-contentalign-left'>
<col class='diff-marker' />
<col class='diff-content' />
<col class='diff-marker' />
<col class='diff-content' />
<tr valign='top'>
<td colspan='2' style="background-color: white; color:black;">← Older revision</td>
<td colspan='2' style="background-color: white; color:black;">Revision as of 18:51, 19 July 2010</td>
</tr><tr><td colspan="2" class="diff-lineno">Line 84:</td>
<td colspan="2" class="diff-lineno">Line 84:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>In July 2008, Ben Martin made a test comparing HW and SW raid with 6 disks. The HW raid was a quite expensive USD 800 Adaptec SAS-31205 PCI Express 12-SATA-port PCI-E x8 hardware RAID card. Compared raid types included [http://www.linux.com/feature/140734|RAID 5, 6 and 10,n2]. Some conclusions: The difference is not big between the expensive HW raid controller and Linux SW RAID. For raid5 Linux was 30 % faster (440 MB/s vs 340 MB/s) for reads. For writes Adaptec swas about 25 % faster (220 MB/s vs 175 MB/s). Keld's remarks: The Adaptec controller actually slowed down disk reading, the single disk read on the motherboard was 90 MB/s while via the Adaptec controller it was only 70 MB/s. Writing was faster, tho. raid10,f2 and raid10,o2 was not included in the test. Ben reported the Adaptec controller to give around 310 MB/s read performance for raid1, while a raid10,f2 would probably have given around 400 MB/s via the Adaptec controller, and around 600 MB/s with the 6 disks on the motherboard SATA controller and a reasonable extra controller, given other benchmarks as noted in this section, giving raid10,f2 a 95 % utilization of the cumulated IO bandwidth. For raid5 the read/write difference could possibly be explained by which chunk size to use, in Linux raid5 reading improves with bigger chunk sizes, while writing degrades.</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>In July 2008, Ben Martin made a test comparing HW and SW raid with 6 disks. The HW raid was a quite expensive USD 800 Adaptec SAS-31205 PCI Express 12-SATA-port PCI-E x8 hardware RAID card. Compared raid types included [http://www.linux.com/feature/140734|RAID 5, 6 and 10,n2]. Some conclusions: The difference is not big between the expensive HW raid controller and Linux SW RAID. For raid5 Linux was 30 % faster (440 MB/s vs 340 MB/s) for reads. For writes Adaptec swas about 25 % faster (220 MB/s vs 175 MB/s). Keld's remarks: The Adaptec controller actually slowed down disk reading, the single disk read on the motherboard was 90 MB/s while via the Adaptec controller it was only 70 MB/s. Writing was faster, tho. raid10,f2 and raid10,o2 was not included in the test. Ben reported the Adaptec controller to give around 310 MB/s read performance for raid1, while a raid10,f2 would probably have given around 400 MB/s via the Adaptec controller, and around 600 MB/s with the 6 disks on the motherboard SATA controller and a reasonable extra controller, given other benchmarks as noted in this section, giving raid10,f2 a 95 % utilization of the cumulated IO bandwidth. For raid5 the read/write difference could possibly be explained by which chunk size to use, in Linux raid5 reading improves with bigger chunk sizes, while writing degrades.</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>In 2009 [http://alephnull.com/benchmarks/sata2009/chunksize.html A Comparison of Chunk Size for Software RAID-5] was done with chunk sizes of 4 KiB to 64 MiB. It was found that chunk sizes of 128 KiB gave the best overall performance. The test was done on a Supermicro AOC-SAT2-MV8 controller with 8 SATA II ports, and connected to a 32-bit PCI slot, which could explain the 130 MB/s max found.  </div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>In 2009 [http://alephnull.com/benchmarks/sata2009/chunksize.html A Comparison of Chunk Size for Software RAID-5] was done <ins class="diffchange diffchange-inline">by Rik Faith </ins>with chunk sizes of 4 KiB to 64 MiB. It was found that chunk sizes of 128 KiB gave the best overall performance. The test was done on a Supermicro AOC-SAT2-MV8 controller with 8 SATA II ports, and connected to a 32-bit PCI slot, which could explain the 130 MB/s max found.  </div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>A benchmark comparing [http://louwrentius.blogspot.com/2010/05/raid-level-and-chunk-size-benchmarks.html chunk sizes] from 4 to 1024 KiB on various RAID types (0, 5, 6, 10) was made in May 2010. For RAID types 5 and 6 it seems like a chunk size of 64 KiB is optimal, while for the other RAID types a chunk size of 512 KiB seemed to give the best results. The test were done on a controller which had an upper limit on about 350 MB/s. It is unclear which layout was used with RAID-10.</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>A benchmark comparing [http://louwrentius.blogspot.com/2010/05/raid-level-and-chunk-size-benchmarks.html chunk sizes] from 4 to 1024 KiB on various RAID types (0, 5, 6, 10) was made in May 2010. For RAID types 5 and 6 it seems like a chunk size of 64 KiB is optimal, while for the other RAID types a chunk size of 512 KiB seemed to give the best results. The test were done on a controller which had an upper limit on about 350 MB/s. It is unclear which layout was used with RAID-10.</div></td></tr>
</table>Keldhttps://raid.wiki.kernel.org/index.php?title=Performance&diff=2027&oldid=prevKeld: /* Other benchmarks from 2007-2010 */2010-07-19T18:41:20Z<p><span dir="auto"><span class="autocomment">Other benchmarks from 2007-2010</span></span></p>
<table class='diff diff-contentalign-left'>
<col class='diff-marker' />
<col class='diff-content' />
<col class='diff-marker' />
<col class='diff-content' />
<tr valign='top'>
<td colspan='2' style="background-color: white; color:black;">← Older revision</td>
<td colspan='2' style="background-color: white; color:black;">Revision as of 18:41, 19 July 2010</td>
</tr><tr><td colspan="2" class="diff-lineno">Line 84:</td>
<td colspan="2" class="diff-lineno">Line 84:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>In July 2008, Ben Martin made a test comparing HW and SW raid with 6 disks. The HW raid was a quite expensive USD 800 Adaptec SAS-31205 PCI Express 12-SATA-port PCI-E x8 hardware RAID card. Compared raid types included [http://www.linux.com/feature/140734|RAID 5, 6 and 10,n2]. Some conclusions: The difference is not big between the expensive HW raid controller and Linux SW RAID. For raid5 Linux was 30 % faster (440 MB/s vs 340 MB/s) for reads. For writes Adaptec swas about 25 % faster (220 MB/s vs 175 MB/s). Keld's remarks: The Adaptec controller actually slowed down disk reading, the single disk read on the motherboard was 90 MB/s while via the Adaptec controller it was only 70 MB/s. Writing was faster, tho. raid10,f2 and raid10,o2 was not included in the test. Ben reported the Adaptec controller to give around 310 MB/s read performance for raid1, while a raid10,f2 would probably have given around 400 MB/s via the Adaptec controller, and around 600 MB/s with the 6 disks on the motherboard SATA controller and a reasonable extra controller, given other benchmarks as noted in this section, giving raid10,f2 a 95 % utilization of the cumulated IO bandwidth. For raid5 the read/write difference could possibly be explained by which chunk size to use, in Linux raid5 reading improves with bigger chunk sizes, while writing degrades.</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>In July 2008, Ben Martin made a test comparing HW and SW raid with 6 disks. The HW raid was a quite expensive USD 800 Adaptec SAS-31205 PCI Express 12-SATA-port PCI-E x8 hardware RAID card. Compared raid types included [http://www.linux.com/feature/140734|RAID 5, 6 and 10,n2]. Some conclusions: The difference is not big between the expensive HW raid controller and Linux SW RAID. For raid5 Linux was 30 % faster (440 MB/s vs 340 MB/s) for reads. For writes Adaptec swas about 25 % faster (220 MB/s vs 175 MB/s). Keld's remarks: The Adaptec controller actually slowed down disk reading, the single disk read on the motherboard was 90 MB/s while via the Adaptec controller it was only 70 MB/s. Writing was faster, tho. raid10,f2 and raid10,o2 was not included in the test. Ben reported the Adaptec controller to give around 310 MB/s read performance for raid1, while a raid10,f2 would probably have given around 400 MB/s via the Adaptec controller, and around 600 MB/s with the 6 disks on the motherboard SATA controller and a reasonable extra controller, given other benchmarks as noted in this section, giving raid10,f2 a 95 % utilization of the cumulated IO bandwidth. For raid5 the read/write difference could possibly be explained by which chunk size to use, in Linux raid5 reading improves with bigger chunk sizes, while writing degrades.</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>A benchmark comparing [http://louwrentius.blogspot.com/2010/05/raid-level-and-chunk-size-benchmarks.html<del class="diffchange diffchange-inline">|</del>chunk sizes] from 4 to 1024 KiB on various RAID types (0, 5, 6, 10) was made in May 2010. For RAID types 5 and 6 it seems like a chunk size of 64 KiB is optimal, while for the other RAID types a chunk size of 512 KiB seemed to give the best results. The test were done on a controller which had an upper limit on about 350 MB/s. It is unclear which layout was used with RAID-10.</div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins class="diffchange diffchange-inline">In 2009 [http://alephnull.com/benchmarks/sata2009/chunksize.html A Comparison of Chunk Size for Software RAID-5] was done with chunk sizes of 4 KiB to 64 MiB. It was found that chunk sizes of 128 KiB gave the best overall performance. The test was done on a Supermicro AOC-SAT2-MV8 controller with 8 SATA II ports, and connected to a 32-bit PCI slot, which could explain the 130 MB/s max found. </ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div> </div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>A benchmark comparing [http://louwrentius.blogspot.com/2010/05/raid-level-and-chunk-size-benchmarks.html chunk sizes] from 4 to 1024 KiB on various RAID types (0, 5, 6, 10) was made in May 2010. For RAID types 5 and 6 it seems like a chunk size of 64 KiB is optimal, while for the other RAID types a chunk size of 512 KiB seemed to give the best results. The test were done on a controller which had an upper limit on about 350 MB/s. It is unclear which layout was used with RAID-10.</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>=Some problem solving for benchmarking=</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>=Some problem solving for benchmarking=</div></td></tr>
</table>Keldhttps://raid.wiki.kernel.org/index.php?title=Performance&diff=2026&oldid=prevKeld: /* Other benchmarks from 2007-2010 */2010-07-19T18:26:36Z<p><span dir="auto"><span class="autocomment">Other benchmarks from 2007-2010</span></span></p>
<table class='diff diff-contentalign-left'>
<col class='diff-marker' />
<col class='diff-content' />
<col class='diff-marker' />
<col class='diff-content' />
<tr valign='top'>
<td colspan='2' style="background-color: white; color:black;">← Older revision</td>
<td colspan='2' style="background-color: white; color:black;">Revision as of 18:26, 19 July 2010</td>
</tr><tr><td colspan="2" class="diff-lineno">Line 49:</td>
<td colspan="2" class="diff-lineno">Line 49:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>My kernel did not allow me to test RAID10,o2 as this is only supported from kernel 2.6.18.</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>My kernel did not allow me to test RAID10,o2 as this is only supported from kernel 2.6.18.</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>=Other benchmarks from 2007-<del class="diffchange diffchange-inline">2008</del>=</div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>=Other benchmarks from 2007-<ins class="diffchange diffchange-inline">2010</ins>=</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>Durval Menezes [http://www.durval.com.br/raid_performance_20081001.txt repeated the above benchmark] in Oct 2008 with a different configuration (3 500GB SATA disks) and a newer kernel (2.6.24), and also included the RAID10,o2 mode, reaching very similar conclusions (nothing beats RAID10,f2 overall).</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>Durval Menezes [http://www.durval.com.br/raid_performance_20081001.txt repeated the above benchmark] in Oct 2008 with a different configuration (3 500GB SATA disks) and a newer kernel (2.6.24), and also included the RAID10,o2 mode, reaching very similar conclusions (nothing beats RAID10,f2 overall).</div></td></tr>
<tr><td colspan="2" class="diff-lineno">Line 82:</td>
<td colspan="2" class="diff-lineno">Line 82:</td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>In July 2008, Jon Nelson made [http://blog.jamponi.net/2008/07/raid56-and-10-benchmarks-on-26255_10.html test of levels 5, 6, and 10, with differing chunk sizes from 64 kiB to 2 MiB] on 4 SATA drives. This was for sequential read and writes on the raw raid types. With a file system, differences in write perfornance would probably smoothen out write differences due to the effect of the IO scheduler (elevator). Results include high performance of raid10,f2 - around 3.80, and high performance of raid5 and rsid6. Especially with bigger chunk sizes 512 kiB - 2 MiB, raid5 obtained maximum 3.44 times the speed of a single drive, and raid6 got a factor 2.96. This is probably due to the even distribution of parity chunks, that means that reads are distributed evenly on all involved (4) drives.</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>In July 2008, Jon Nelson made [http://blog.jamponi.net/2008/07/raid56-and-10-benchmarks-on-26255_10.html test of levels 5, 6, and 10, with differing chunk sizes from 64 kiB to 2 MiB] on 4 SATA drives. This was for sequential read and writes on the raw raid types. With a file system, differences in write perfornance would probably smoothen out write differences due to the effect of the IO scheduler (elevator). Results include high performance of raid10,f2 - around 3.80, and high performance of raid5 and rsid6. Especially with bigger chunk sizes 512 kiB - 2 MiB, raid5 obtained maximum 3.44 times the speed of a single drive, and raid6 got a factor 2.96. This is probably due to the even distribution of parity chunks, that means that reads are distributed evenly on all involved (4) drives.</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'>−</td><td style="background: #ffa; color:black; font-size: smaller;"><div>In July 2008, Ben Martin made a test comparing HW and SW raid with 6 disks. The HW raid was a quite expensive USD 800 Adaptec SAS-31205 PCI Express 12-SATA-port PCI-E x8 hardware RAID card. Compared raid types included [http://www.linux.com/feature/140734|RAID 5, 6 and 10,n2]. Some conclusions: The difference is not big between the expensive HW raid controller and Linux SW RAID. For raid5 Linux was 30 % faster (440 MB/s vs 340 MB/s) for reads. For writes Adaptec swas about 25 % faster (220 MB/s vs 175 MB/s). Keld's remarks: The Adaptec controller actually slowed down disk reading, the single disk read on the motherboard was 90 MB/s while via the Adaptec controller it was only 70 MB/s. Writing was faster, tho. raid10,f2 and raid10,o2 was not included in the test. Ben reported the Adaptec controller to give around 310 MB/s read performance for raid1, while a raid10,f2 would probably have given around 400 MB/s via the Adaptec controller, and around 600 MB/s with the 6 disks on the motherboard SATA controller and a reasonable extra <del class="diffchange diffchange-inline">contoller</del>, given other benchmarks as noted in this section, giving raid10,f2 a 95 % utilization of the cumulated IO bandwidth. For raid5 the read/write difference could possibly be explained by which chunk size to use, in Linux raid5 reading improves with bigger chunk sizes, while writing degrades.</div></td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div>In July 2008, Ben Martin made a test comparing HW and SW raid with 6 disks. The HW raid was a quite expensive USD 800 Adaptec SAS-31205 PCI Express 12-SATA-port PCI-E x8 hardware RAID card. Compared raid types included [http://www.linux.com/feature/140734|RAID 5, 6 and 10,n2]. Some conclusions: The difference is not big between the expensive HW raid controller and Linux SW RAID. For raid5 Linux was 30 % faster (440 MB/s vs 340 MB/s) for reads. For writes Adaptec swas about 25 % faster (220 MB/s vs 175 MB/s). Keld's remarks: The Adaptec controller actually slowed down disk reading, the single disk read on the motherboard was 90 MB/s while via the Adaptec controller it was only 70 MB/s. Writing was faster, tho. raid10,f2 and raid10,o2 was not included in the test. Ben reported the Adaptec controller to give around 310 MB/s read performance for raid1, while a raid10,f2 would probably have given around 400 MB/s via the Adaptec controller, and around 600 MB/s with the 6 disks on the motherboard SATA controller and a reasonable extra <ins class="diffchange diffchange-inline">controller</ins>, given other benchmarks as noted in this section, giving raid10,f2 a 95 % utilization of the cumulated IO bandwidth. For raid5 the read/write difference could possibly be explained by which chunk size to use, in Linux raid5 reading improves with bigger chunk sizes, while writing degrades<ins class="diffchange diffchange-inline">.</ins></div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div> </div></td></tr>
<tr><td colspan="2"> </td><td class='diff-marker'>+</td><td style="background: #cfc; color:black; font-size: smaller;"><div><ins class="diffchange diffchange-inline">A benchmark comparing [http://louwrentius.blogspot.com/2010/05/raid-level-and-chunk-size-benchmarks.html|chunk sizes] from 4 to 1024 KiB on various RAID types (0, 5, 6, 10) was made in May 2010. For RAID types 5 and 6 it seems like a chunk size of 64 KiB is optimal, while for the other RAID types a chunk size of 512 KiB seemed to give the best results. The test were done on a controller which had an upper limit on about 350 MB/s. It is unclear which layout was used with RAID-10</ins>.</div></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"></td></tr>
<tr><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>=Some problem solving for benchmarking=</div></td><td class='diff-marker'> </td><td style="background: #eee; color:black; font-size: smaller;"><div>=Some problem solving for benchmarking=</div></td></tr>
</table>Keld