File:Graid-performance.png
(add raw data for Linux RAID4) |
|||
| Line 15: | Line 15: | ||
(using dd if=/dev/random bs=16m count=200). | (using dd if=/dev/random bs=16m count=200). | ||
| − | For these tests, read-ahead cache value was GREATLY increased from 6.2-RELEASE's default of 8: '''sysctl -w vfs.read_max=128'''. Linux's read-ahead cache value was similarly greatly increased from 256 to 4096 using '''hdparm /dev/md0 -a4192'''. I noticed Linux automatically bumped the default 256 value up to 1024 for the RAID5 array, so I retested it and discovered the "sweet spot" for a 5-member Linux RAID5 array to be 8192. | + | For these tests, read-ahead cache value was GREATLY increased from 6.2-RELEASE's default of 8: '''sysctl -w vfs.read_max=128'''. Linux's read-ahead cache value was similarly greatly increased from 256 to 4096 using '''hdparm /dev/md0 -a4192'''. I noticed Linux automatically bumped the default 256 value up to 1024 for the RAID5 array, so I retested it and discovered the "sweet spot" for a 5-member Linux RAID5 or RAID4 array to be 8192. |
The "-r" option (use parity drive on reads) for graid3 was also tested but typically resulted in slight to severe performance decrease, so was not included on the graph, although its raw data is available below. | The "-r" option (use parity drive on reads) for graid3 was also tested but typically resulted in slight to severe performance decrease, so was not included on the graph, although its raw data is available below. | ||
| Line 21: | Line 21: | ||
Raw data follows: | Raw data follows: | ||
| + | Linux RAID4: 3x250/2x500 200 168 160 168 162 * Hdparm -a 8192 | ||
Linux RAID5: 3x250/2x500 161 173 163 164 173 * Hdparm -a 8192 | Linux RAID5: 3x250/2x500 161 173 163 164 173 * Hdparm -a 8192 | ||
Linux RAID1: 3x250/2x500 68 152 213 251 258 * Hdparm -a 4096 | Linux RAID1: 3x250/2x500 68 152 213 251 258 * Hdparm -a 4096 | ||
Latest revision as of 02:55, 29 December 2007
Chart showing performance of graid3 and gmirror (and some proprietary RAID1 hardware implementations).
System:
FreeBSD OS: amd64 6.2-RELEASE
Linux OS: amd64 Ubuntu Server 7.04
Athlon X2 5000+, 2GB RAM
3x Western Digital 250GB drives (WDC WD2500JS-22NCB1 10.02E02 SATA-300)
2x Western Digital 500GB drives (WDC WD5000AAKS-00YGA0 12.01C02)
Nvidia nForce onboard RAID controller
Promise TX2300 RAID controller
Each process shown in the graph is a cp to /dev/null of an individually generated 3200MB random binary file
(using dd if=/dev/random bs=16m count=200).
For these tests, read-ahead cache value was GREATLY increased from 6.2-RELEASE's default of 8: sysctl -w vfs.read_max=128. Linux's read-ahead cache value was similarly greatly increased from 256 to 4096 using hdparm /dev/md0 -a4192. I noticed Linux automatically bumped the default 256 value up to 1024 for the RAID5 array, so I retested it and discovered the "sweet spot" for a 5-member Linux RAID5 or RAID4 array to be 8192.
The "-r" option (use parity drive on reads) for graid3 was also tested but typically resulted in slight to severe performance decrease, so was not included on the graph, although its raw data is available below.
Raw data follows:
Linux RAID4: 3x250/2x500 200 168 160 168 162 * Hdparm -a 8192 Linux RAID5: 3x250/2x500 161 173 163 164 173 * Hdparm -a 8192 Linux RAID1: 3x250/2x500 68 152 213 251 258 * Hdparm -a 4096 Nvidia onboard RAID0+1: 3x250/1x500 97 142 160 151 144 * 7s delta on 2-process copy... 66 second delta on 5-process copy Graid3: 3x250/2x500 213 137 106 105 105 Graid3: 3x250/2x500 -r 187 144 103 105 107 Linux RAID1: 3x250 62 123 155 158 162 * Hdparm -a 4096 Graid3: 3x250 115 98 77 76 75 Graid3: 3x250 -r 116 88 79 76 75 Linux RAID1: 2x250 62 114 108 102 110 * 31s delta on 4-copy... 30-sec delta on 5-copy Nvidia onboard RAID1: 2x250 60 108 90 96 100 * 15 secs delta on 2-process copy / 38 seconds on 5-process! Baseline: 1x500 76 70 72 72 72 Gmirror: 3x250/2x500 round-robin 63 73 69 71 72 Promise TX-2300 RAID1: 2x250 61 52 54 55 57 * 15 sec delta on 5-process copy Gmirror: 3x250 round-robin 59 53 53 56 55 Gmirror: 2x250 round-robin 58 49 51 51 50 Baseline: 1x250 56 56 55 55 56 Gmirror: 3x250 split 128k 55 46 49 55 54 Gmirror: 2x250 split 128k 55 47 47 47 48 Gmirror: 3x250/2x500 split 128k 54 50 47 47 47
"Delta" in raw numbers refers to individual processes finishing significantly faster than others, though all were started simultaneously.
File history
Click on a date/time to view the file as it appeared at that time.
| Date/Time | Thumbnail | Dimensions | User | Comment | |
|---|---|---|---|---|---|
| current | 03:28, 29 December 2007 |  | 650 × 400 (25 KB) | Jimbo (Talk | contribs) | |
| 02:54, 29 December 2007 | No thumbnail | (25 KB) | Jimbo (Talk | contribs) | (add Linux RAID4 data) | |
| 01:13, 28 December 2007 |  | 650 × 400 (22 KB) | Jimbo (Talk | contribs) | (Linux RAID5 data added) | |
| 23:30, 27 December 2007 |  | 650 × 400 (21 KB) | Jimbo (Talk | contribs) | (added some linux and nvidia 0+1 data) | |
| 21:00, 26 December 2007 |  | 650 × 400 (22 KB) | Jimbo (Talk | contribs) |
- Edit this file using an external application (See the setup instructions for more information)
File usage
The following page links to this file:
