opentoe Posted January 19, 2011 Share Posted January 19, 2011 All drives are Western Digital Black 1TB drives /dev/sda: Timing cached reads: 9218 MB in 2.00 seconds = 4616.89 MB/sec Timing buffered disk reads: 310 MB in 3.01 seconds = 102.91 MB/sec /dev/sdb: Timing cached reads: 9052 MB in 2.00 seconds = 4533.87 MB/sec Timing buffered disk reads: 308 MB in 3.02 seconds = 102.04 MB/sec /dev/sdc: Timing cached reads: 9128 MB in 2.00 seconds = 4572.11 MB/sec Timing buffered disk reads: 328 MB in 3.00 seconds = 109.21 MB/sec /dev/sdd: Timing cached reads: 9144 MB in 2.00 seconds = 4580.01 MB/sec Timing buffered disk reads: 312 MB in 3.01 seconds = 103.76 MB/sec /dev/sde: Timing cached reads: 8962 MB in 2.00 seconds = 4488.32 MB/sec Timing buffered disk reads: 332 MB in 3.01 seconds = 110.39 MB/sec /dev/sdf: Timing cached reads: 8698 MB in 2.00 seconds = 4355.76 MB/sec Timing buffered disk reads: 312 MB in 3.00 seconds = 103.96 MB/sec /dev/sdg: Timing cached reads: 8228 MB in 2.00 seconds = 4119.81 MB/sec Timing buffered disk reads: 56 MB in 3.03 seconds = 18.51 MB/sec Quote Link to comment
dgaschk Posted January 19, 2011 Share Posted January 19, 2011 What type of drives are these? Quote Link to comment
opentoe Posted January 19, 2011 Author Share Posted January 19, 2011 Sorry, that would have been a good thing to include. They all are Western Digital Black 1TB drives. Quote Link to comment
dgaschk Posted January 19, 2011 Share Posted January 19, 2011 What command did you use? I think the last one is a lemon. Quote Link to comment
BRiT Posted January 19, 2011 Share Posted January 19, 2011 I think a more meaningful test than 'hdparm -tT /dev/[sh]d?' is needed to get any sort of real-world performance ideas. It was proposed earlier and elsewhere on these forums that something like the following 'diskspeed.sh' should be suitable. diskspeed.sh /mnt/disk0/testspeed diskspeed.sh /mnt/disk1/testspeed diskspeed.sh /mnt/disk2/testspeed #!/bin/bash if [ ! -z "${1}" ] then TMPFILE="${1}" else TMPFILE="test.$$" fi trap "rm -f ${TMPFILE}" EXIT HUP INT QUIT TERM echo "`date` writing to: ${TMPFILE}" dd if=/dev/zero of="${TMPFILE}" count=8000000 bs=1024 echo "`date` Done." ls -l --si ${TMPFILE} rm -f "${TMPFILE}" Quote Link to comment
dgaschk Posted January 19, 2011 Share Posted January 19, 2011 root@192:~# hdparm -tT /dev/[sh]d? Western Digital Caviar SE family /dev/hdc: Timing cached reads: 2672 MB in 2.00 seconds = 1336.33 MB/sec Timing buffered disk reads: 164 MB in 3.00 seconds = 54.62 MB/sec Maxtor DiamondMax 10 family (ATA/133 and SATA/150) /dev/hdd: Timing cached reads: 2706 MB in 2.00 seconds = 1352.54 MB/sec Timing buffered disk reads: 190 MB in 3.02 seconds = 62.98 MB/sec Seagate (ST32000542AS) /dev/sda: Timing cached reads: 2670 MB in 2.00 seconds = 1335.25 MB/sec Timing buffered disk reads: 284 MB in 3.00 seconds = 94.63 MB/sec WD (WD20EARS) /dev/sdb: Timing cached reads: 2672 MB in 2.00 seconds = 1335.58 MB/sec Timing buffered disk reads: 318 MB in 3.00 seconds = 105.93 MB/sec Hitachi Deskstar 7K2000 /dev/sdc: Timing cached reads: 2680 MB in 2.00 seconds = 1339.98 MB/sec Timing buffered disk reads: 342 MB in 3.01 seconds = 113.66 MB/sec Seagate (ST32000542AS) /dev/sdd: Timing cached reads: 2698 MB in 2.00 seconds = 1349.02 MB/sec Timing buffered disk reads: 328 MB in 3.01 seconds = 108.82 MB/sec Is this the flash drive? /dev/sde: Timing cached reads: 2592 MB in 2.00 seconds = 1296.19 MB/sec Timing buffered disk reads: 54 MB in 3.03 seconds = 17.84 MB/sec root@192:~# hdparm -tT /dev/sde /dev/sde: Timing cached reads: 2556 MB in 2.00 seconds = 1277.70 MB/sec Timing buffered disk reads: 54 MB in 3.03 seconds = 17.84 MB/sec root@192:~# I will post results of diskspeed.sh Quote Link to comment
kaiguy Posted January 20, 2011 Share Posted January 20, 2011 Well this is weird. My WD Caviar Black drive is apparently slower than my WD20EARS drives. THAT is interesting. /dev/sda: SanDisk Cruzer Micro 2GB Timing cached reads: 10370 MB in 2.00 seconds = 5191.64 MB/sec Timing buffered disk reads: 66 MB in 3.05 seconds = 21.61 MB/sec /dev/sdb: Seagate Barracuda LP 2 TB Timing cached reads: 10352 MB in 2.00 seconds = 5182.66 MB/sec Timing buffered disk reads: 344 MB in 3.01 seconds = 114.18 MB/sec /dev/sdc: WD15EARS Timing cached reads: 9918 MB in 2.00 seconds = 4964.76 MB/sec Timing buffered disk reads: 280 MB in 3.00 seconds = 93.21 MB/sec /dev/sdd: WD15EARS Timing cached reads: 9972 MB in 2.00 seconds = 4992.14 MB/sec Timing buffered disk reads: 272 MB in 3.02 seconds = 90.00 MB/sec /dev/sde: WD20EARS Timing cached reads: 9484 MB in 2.00 seconds = 4747.51 MB/sec Timing buffered disk reads: 346 MB in 3.00 seconds = 115.18 MB/sec /dev/sdf: WD20EARS Timing cached reads: 10368 MB in 2.00 seconds = 5190.15 MB/sec Timing buffered disk reads: 354 MB in 3.01 seconds = 117.43 MB/sec /dev/sdg: WD20EARS Timing cached reads: 9846 MB in 2.00 seconds = 4928.60 MB/sec Timing buffered disk reads: 340 MB in 3.01 seconds = 112.81 MB/sec /dev/sdh: WD Caviar Black 1 TB Timing cached reads: 10154 MB in 2.00 seconds = 5082.47 MB/sec Timing buffered disk reads: 318 MB in 3.00 seconds = 105.97 MB/sec Quote Link to comment
lionelhutz Posted January 20, 2011 Share Posted January 20, 2011 Any reason why?? Peter Quote Link to comment
BRiT Posted January 20, 2011 Share Posted January 20, 2011 Well this is weird. My WD Caviar Black drive is apparently slower than my WD20EARS drives. THAT is interesting. That's why in previous discussions this forum has suggested the hdparm tests are nearly meaningless and why it was mentioned using a test with real world aspects such as diskspeed.sh. Quote Link to comment
opentoe Posted January 20, 2011 Author Share Posted January 20, 2011 I think a more meaningful test than 'hdparm -tT /dev/[sh]d?' is needed to get any sort of real-world performance ideas. It was proposed earlier and elsewhere on these forums that something like the following 'diskspeed.sh' should be suitable. diskspeed.sh /mnt/disk0/testspeed diskspeed.sh /mnt/disk1/testspeed diskspeed.sh /mnt/disk2/testspeed #!/bin/bash if [ ! -z "${1}" ] then TMPFILE="${1}" else TMPFILE="test.$$" fi trap "rm -f ${TMPFILE}" EXIT HUP INT QUIT TERM echo "`date` writing to: ${TMPFILE}" dd if=/dev/zero of="${TMPFILE}" count=8000000 bs=1024 echo "`date` Done." ls -l --si ${TMPFILE} rm -f "${TMPFILE}" I copied the code into a file and named it diskspeed.sh and then ran it in my shell. Came back with this: diskspeed /mnt/disk1/testspeed -bash: ./diskspeed.sh: /bin/bash^M: bad interpreter: No such file or directory Quote Link to comment
WeeboTech Posted January 20, 2011 Share Posted January 20, 2011 I copied the code into a file and named it diskspeed.sh and then ran it in my shell. Came back with this: diskspeed /mnt/disk1/testspeed -bash: ./diskspeed.sh: /bin/bash^M: bad interpreter: No such file or directory It may not be unix LF terminated. You can try and convert it on your box with. fromdos < ./diskspeed.sh > ./diskspeed.bash mv ./diskspeed.bash ./diskspeed.sh chmod u+rwx ./diskspeed.sh Quote Link to comment
WeeboTech Posted January 20, 2011 Share Posted January 20, 2011 My newer utility shows the status over time. I do a 10GB file to make sure all 8GB is flushed. For the record I have my system tweaked for max/IO and caching. Therefore smaller files burst onto the drive lightning fast. Seagate - 1.5TB 7200RPM 32MB cache empty fs, ST31500341AS 463955968 bytes (464 MB) copied, 5.01395 s, 92.5 MB/s 689383424 bytes (689 MB) copied, 10.0301 s, 68.7 MB/s 905988096 bytes (906 MB) copied, 15.0806 s, 60.1 MB/s 1174423552 bytes (1.2 GB) copied, 20.1146 s, 58.4 MB/s 1425033216 bytes (1.4 GB) copied, 25.0855 s, 56.8 MB/s 1657558016 bytes (1.7 GB) copied, 30.099 s, 55.1 MB/s ..... 10240000000 bytes (10 GB) copied, 223.783 s, 45.8 MB/s uncomment the reads below to get the read speed tests. Note syncs are called, so this could cause all your drives to spin up. You may want to spin them up manually before the testing or comment out the syncs. #!/bin/bash if [ -z "$1" ] then echo "Usage: $0 outputfilename" exit fi if [ -f "$1" ] then echo "removing: $1" rm -vf $1 sync fi bs=1024 count=4000000 count=10000000 total=$(( $bs * $count)) echo "writing $total bytes to: $1" touch $1;rm -f $1 dd if=/dev/zero bs=$bs count=$count of=$1 & BGPID=$! trap "kill $BGPID; rm -vf '$1'; exit" INT HUP QUIT TERM EXIT sleep 5 while ps --no-heading -fp $BGPID >/dev/null do kill -USR1 $BGPID sleep 5 done trap "rm -vf '$1'; exit" INT HUP QUIT TERM EXIT echo "write complete, syncing" # sync # echo "reading from: $1" # dd if=$1 bs=$bs count=$count of=/dev/null echo "removing: $1" rm -vf $1 Quote Link to comment
BRiT Posted January 20, 2011 Share Posted January 20, 2011 My newer utility shows the status over time. I do a 10GB file to make sure all 8GB is flushed. Thanks for sharing! It's certainly more interesting to see the speeds over time as opposed to just a final number. Quote Link to comment
BRiT Posted January 20, 2011 Share Posted January 20, 2011 Here's full results from my array on a Slackware 13.2 (Current) distro with unRAID 5.0 beta 2 on Linux 2.6.36.3, 4GB DDR3 RAM, i3 530 CPU, MSI H55 miniATX motherboard with all drives connected to the motherboard SATA ports. No tweaks were done for the kernel config values. Disk 2 always benches the slowest out of the three drives. Overall I'm very pleased with the performance of my green system. I forgot to stop cache_dirs from running, so it's likely to have impacted the tests somewhat. Parity = 2TB Seagate 5900RPM ST32000542AS CC35 [/dev/sdb] Drive1 = 2TB WD 5400RPM WD20EADS-00R6B0 57% full [/dev/sdc] Drive2 = 2TB WD 5400RPM WD20EADS-00R6B0 52% full [/dev/sdd] Drive3 = 2TB WD 5400RPM WD20EADS-00R6B0 27% full [/dev/sde] writing 10240000000 bytes to: /mnt/disk1/Media/speedtest 1182299136 bytes (1.2 GB) copied, 5.06237 s, 234 MB/s 1339849728 bytes (1.3 GB) copied, 10.0975 s, 133 MB/s 1479283712 bytes (1.5 GB) copied, 15.3355 s, 96.5 MB/s 1628444672 bytes (1.6 GB) copied, 20.1253 s, 80.9 MB/s 1779318784 bytes (1.8 GB) copied, 25.2707 s, 70.4 MB/s 1939300352 bytes (1.9 GB) copied, 30.172 s, 64.3 MB/s 2098971648 bytes (2.1 GB) copied, 35.3808 s, 59.3 MB/s 2269127680 bytes (2.3 GB) copied, 40.3962 s, 56.2 MB/s 2435516416 bytes (2.4 GB) copied, 45.4433 s, 53.6 MB/s 2541936640 bytes (2.5 GB) copied, 50.2263 s, 50.6 MB/s 2701647872 bytes (2.7 GB) copied, 55.2974 s, 48.9 MB/s 2857384960 bytes (2.9 GB) copied, 60.2835 s, 47.4 MB/s 3026468864 bytes (3.0 GB) copied, 65.3262 s, 46.3 MB/s 3191779328 bytes (3.2 GB) copied, 70.3372 s, 45.4 MB/s 3362237440 bytes (3.4 GB) copied, 75.4915 s, 44.5 MB/s 3477083136 bytes (3.5 GB) copied, 80.4437 s, 43.2 MB/s 3637737472 bytes (3.6 GB) copied, 85.5002 s, 42.5 MB/s 3793139712 bytes (3.8 GB) copied, 90.4934 s, 41.9 MB/s 3944113152 bytes (3.9 GB) copied, 95.3878 s, 41.3 MB/s 4117800960 bytes (4.1 GB) copied, 100.684 s, 40.9 MB/s 4274891776 bytes (4.3 GB) copied, 105.418 s, 40.6 MB/s 4448112640 bytes (4.4 GB) copied, 110.432 s, 40.3 MB/s 4580073472 bytes (4.6 GB) copied, 115.71 s, 39.6 MB/s 4721827840 bytes (4.7 GB) copied, 120.587 s, 39.2 MB/s 4876427264 bytes (4.9 GB) copied, 125.476 s, 38.9 MB/s 5036489728 bytes (5.0 GB) copied, 130.651 s, 38.5 MB/s 5189419008 bytes (5.2 GB) copied, 135.61 s, 38.3 MB/s 5358154752 bytes (5.4 GB) copied, 140.52 s, 38.1 MB/s 5518832640 bytes (5.5 GB) copied, 145.604 s, 37.9 MB/s 5653817344 bytes (5.7 GB) copied, 150.569 s, 37.5 MB/s 5813190656 bytes (5.8 GB) copied, 155.688 s, 37.3 MB/s 5962961920 bytes (6.0 GB) copied, 160.577 s, 37.1 MB/s 6114904064 bytes (6.1 GB) copied, 165.662 s, 36.9 MB/s 6281909248 bytes (6.3 GB) copied, 170.783 s, 36.8 MB/s 6449206272 bytes (6.4 GB) copied, 175.87 s, 36.7 MB/s 6610857984 bytes (6.6 GB) copied, 180.922 s, 36.5 MB/s 6748619776 bytes (6.7 GB) copied, 185.781 s, 36.3 MB/s 6918723584 bytes (6.9 GB) copied, 190.663 s, 36.3 MB/s 7074656256 bytes (7.1 GB) copied, 195.677 s, 36.2 MB/s 7237877760 bytes (7.2 GB) copied, 200.892 s, 36.0 MB/s 7404745728 bytes (7.4 GB) copied, 205.762 s, 36.0 MB/s 7578809344 bytes (7.6 GB) copied, 210.712 s, 36.0 MB/s 7735047168 bytes (7.7 GB) copied, 215.726 s, 35.9 MB/s 7915287552 bytes (7.9 GB) copied, 220.743 s, 35.9 MB/s 8038663168 bytes (8.0 GB) copied, 225.756 s, 35.6 MB/s 8190749696 bytes (8.2 GB) copied, 230.775 s, 35.5 MB/s 8353758208 bytes (8.4 GB) copied, 235.889 s, 35.4 MB/s 8512749568 bytes (8.5 GB) copied, 240.796 s, 35.4 MB/s 8675278848 bytes (8.7 GB) copied, 245.812 s, 35.3 MB/s 8845859840 bytes (8.8 GB) copied, 250.912 s, 35.3 MB/s 9022160896 bytes (9.0 GB) copied, 255.881 s, 35.3 MB/s 9155391488 bytes (9.2 GB) copied, 260.853 s, 35.1 MB/s 9311899648 bytes (9.3 GB) copied, 266.056 s, 35.0 MB/s 9465107456 bytes (9.5 GB) copied, 270.94 s, 34.9 MB/s 9624185856 bytes (9.6 GB) copied, 275.919 s, 34.9 MB/s 9785817088 bytes (9.8 GB) copied, 280.912 s, 34.8 MB/s 9953438720 bytes (10 GB) copied, 285.938 s, 34.8 MB/s 10125725696 bytes (10 GB) copied, 291.058 s, 34.8 MB/s 10240000000 bytes (10 GB) copied, 294.04 s, 34.8 MB/s write complete, syncing reading from: /mnt/disk1/Media/speedtest 10240000000 bytes (10 GB) copied, 127.131 s, 80.5 MB/s removing: /mnt/disk1/Media/speedtest removed `/mnt/disk1/Media/speedtest' writing 10240000000 bytes to: /mnt/disk2/Media/speedtest2 1124855808 bytes (1.1 GB) copied, 5.03711 s, 223 MB/s 1283752960 bytes (1.3 GB) copied, 10.0502 s, 128 MB/s 1449223168 bytes (1.4 GB) copied, 15.0633 s, 96.2 MB/s 1609248768 bytes (1.6 GB) copied, 20.0767 s, 80.2 MB/s 1763769344 bytes (1.8 GB) copied, 25.0909 s, 70.3 MB/s 1928361984 bytes (1.9 GB) copied, 30.1615 s, 63.9 MB/s 2080734208 bytes (2.1 GB) copied, 35.1181 s, 59.2 MB/s 2205663232 bytes (2.2 GB) copied, 40.146 s, 54.9 MB/s 2363275264 bytes (2.4 GB) copied, 45.2642 s, 52.2 MB/s 2518168576 bytes (2.5 GB) copied, 50.1606 s, 50.2 MB/s 2677704704 bytes (2.7 GB) copied, 55.4887 s, 48.3 MB/s 2838207488 bytes (2.8 GB) copied, 60.2491 s, 47.1 MB/s 2997885952 bytes (3.0 GB) copied, 65.364 s, 45.9 MB/s 3138349056 bytes (3.1 GB) copied, 70.2475 s, 44.7 MB/s 3286678528 bytes (3.3 GB) copied, 75.261 s, 43.7 MB/s 3443478528 bytes (3.4 GB) copied, 80.2748 s, 42.9 MB/s 3599113216 bytes (3.6 GB) copied, 85.3369 s, 42.2 MB/s 3754768384 bytes (3.8 GB) copied, 90.3022 s, 41.6 MB/s 3905651712 bytes (3.9 GB) copied, 95.3905 s, 40.9 MB/s 4073230336 bytes (4.1 GB) copied, 100.419 s, 40.6 MB/s 4215989248 bytes (4.2 GB) copied, 105.473 s, 40.0 MB/s 4356403200 bytes (4.4 GB) copied, 110.562 s, 39.4 MB/s 4504107008 bytes (4.5 GB) copied, 115.37 s, 39.0 MB/s 4662859776 bytes (4.7 GB) copied, 120.384 s, 38.7 MB/s 4829432832 bytes (4.8 GB) copied, 125.599 s, 38.5 MB/s 4983738368 bytes (5.0 GB) copied, 130.482 s, 38.2 MB/s 5151731712 bytes (5.2 GB) copied, 135.623 s, 38.0 MB/s 5215300608 bytes (5.2 GB) copied, 140.501 s, 37.1 MB/s 5257125888 bytes (5.3 GB) copied, 145.6 s, 36.1 MB/s 5291166720 bytes (5.3 GB) copied, 151.014 s, 35.0 MB/s 5319533568 bytes (5.3 GB) copied, 158.096 s, 33.6 MB/s 5333222400 bytes (5.3 GB) copied, 160.658 s, 33.2 MB/s 5371667456 bytes (5.4 GB) copied, 166.318 s, 32.3 MB/s 5474563072 bytes (5.5 GB) copied, 170.892 s, 32.0 MB/s 5593273344 bytes (5.6 GB) copied, 175.818 s, 31.8 MB/s 5712490496 bytes (5.7 GB) copied, 181.232 s, 31.5 MB/s 5848003584 bytes (5.8 GB) copied, 185.739 s, 31.5 MB/s 5984182272 bytes (6.0 GB) copied, 191.126 s, 31.3 MB/s 6127165440 bytes (6.1 GB) copied, 195.767 s, 31.3 MB/s 6257838080 bytes (6.3 GB) copied, 200.774 s, 31.2 MB/s 6366057472 bytes (6.4 GB) copied, 205.787 s, 30.9 MB/s 6501087232 bytes (6.5 GB) copied, 210.801 s, 30.8 MB/s 6633485312 bytes (6.6 GB) copied, 215.912 s, 30.7 MB/s 6775339008 bytes (6.8 GB) copied, 220.83 s, 30.7 MB/s 6933488640 bytes (6.9 GB) copied, 225.843 s, 30.7 MB/s 7088283648 bytes (7.1 GB) copied, 230.972 s, 30.7 MB/s 7228150784 bytes (7.2 GB) copied, 236.065 s, 30.6 MB/s 7382291456 bytes (7.4 GB) copied, 241.87 s, 30.5 MB/s 7509073920 bytes (7.5 GB) copied, 246.037 s, 30.5 MB/s 7644255232 bytes (7.6 GB) copied, 251.031 s, 30.5 MB/s 7774278656 bytes (7.8 GB) copied, 256.044 s, 30.4 MB/s 7921079296 bytes (7.9 GB) copied, 261.058 s, 30.3 MB/s 8061321216 bytes (8.1 GB) copied, 266.152 s, 30.3 MB/s 8206024704 bytes (8.2 GB) copied, 271.088 s, 30.3 MB/s 8349679616 bytes (8.3 GB) copied, 276.1 s, 30.2 MB/s 8489257984 bytes (8.5 GB) copied, 281.119 s, 30.2 MB/s 8643121152 bytes (8.6 GB) copied, 287.003 s, 30.1 MB/s 8759255040 bytes (8.8 GB) copied, 291.438 s, 30.1 MB/s 8907309056 bytes (8.9 GB) copied, 296.155 s, 30.1 MB/s 9033146368 bytes (9.0 GB) copied, 301.182 s, 30.0 MB/s 9177702400 bytes (9.2 GB) copied, 306.188 s, 30.0 MB/s 9323189248 bytes (9.3 GB) copied, 311.275 s, 30.0 MB/s 9471884288 bytes (9.5 GB) copied, 316.212 s, 30.0 MB/s 9616405504 bytes (9.6 GB) copied, 321.234 s, 29.9 MB/s 9783323648 bytes (9.8 GB) copied, 326.273 s, 30.0 MB/s 9913463808 bytes (9.9 GB) copied, 331.265 s, 29.9 MB/s 10065272832 bytes (10 GB) copied, 336.278 s, 29.9 MB/s 10220843008 bytes (10 GB) copied, 341.276 s, 29.9 MB/s 10240000000 bytes (10 GB) copied, 341.784 s, 30.0 MB/s write complete, syncing reading from: /mnt/disk2/Media/speedtest2 10240000000 bytes (10 GB) copied, 126.593 s, 80.9 MB/s removing: /mnt/disk2/Media/speedtest2 removed `/mnt/disk2/Media/speedtest2' writing 10240000000 bytes to: /mnt/disk3/Movies/speedtest3 1087149056 bytes (1.1 GB) copied, 5.01133 s, 217 MB/s 1291424768 bytes (1.3 GB) copied, 10.0799 s, 128 MB/s 1440610304 bytes (1.4 GB) copied, 15.0373 s, 95.8 MB/s 1608355840 bytes (1.6 GB) copied, 20.0959 s, 80.0 MB/s 1762226176 bytes (1.8 GB) copied, 25.1452 s, 70.1 MB/s 1928205312 bytes (1.9 GB) copied, 30.8779 s, 62.4 MB/s 2040886272 bytes (2.0 GB) copied, 35.0988 s, 58.1 MB/s 2206102528 bytes (2.2 GB) copied, 40.2014 s, 54.9 MB/s 2369123328 bytes (2.4 GB) copied, 45.1201 s, 52.5 MB/s 2540965888 bytes (2.5 GB) copied, 50.2037 s, 50.6 MB/s 2707919872 bytes (2.7 GB) copied, 55.4583 s, 48.8 MB/s 2857055232 bytes (2.9 GB) copied, 60.2884 s, 47.4 MB/s 3021383680 bytes (3.0 GB) copied, 65.5099 s, 46.1 MB/s 3129422848 bytes (3.1 GB) copied, 70.3087 s, 44.5 MB/s 3298770944 bytes (3.3 GB) copied, 75.3943 s, 43.8 MB/s 3463404544 bytes (3.5 GB) copied, 80.3394 s, 43.1 MB/s 3624467456 bytes (3.6 GB) copied, 85.4665 s, 42.4 MB/s 3784739840 bytes (3.8 GB) copied, 90.3647 s, 41.9 MB/s 3944497152 bytes (3.9 GB) copied, 95.6429 s, 41.2 MB/s 4079123456 bytes (4.1 GB) copied, 100.81 s, 40.5 MB/s 4213462016 bytes (4.2 GB) copied, 105.44 s, 40.0 MB/s 4367881216 bytes (4.4 GB) copied, 110.522 s, 39.5 MB/s 4528706560 bytes (4.5 GB) copied, 115.557 s, 39.2 MB/s 4691792896 bytes (4.7 GB) copied, 120.446 s, 39.0 MB/s 4854633472 bytes (4.9 GB) copied, 125.59 s, 38.7 MB/s 5019960320 bytes (5.0 GB) copied, 130.471 s, 38.5 MB/s 5146626048 bytes (5.1 GB) copied, 135.684 s, 37.9 MB/s 5261125632 bytes (5.3 GB) copied, 140.768 s, 37.4 MB/s 5345543168 bytes (5.3 GB) copied, 145.528 s, 36.7 MB/s 5398018048 bytes (5.4 GB) copied, 150.602 s, 35.8 MB/s 5428945920 bytes (5.4 GB) copied, 155.798 s, 34.8 MB/s 5488870400 bytes (5.5 GB) copied, 161.222 s, 34.0 MB/s 5625488384 bytes (5.6 GB) copied, 165.748 s, 33.9 MB/s 5776375808 bytes (5.8 GB) copied, 170.813 s, 33.8 MB/s 5907337216 bytes (5.9 GB) copied, 175.823 s, 33.6 MB/s 6049189888 bytes (6.0 GB) copied, 180.72 s, 33.5 MB/s 6176223232 bytes (6.2 GB) copied, 185.935 s, 33.2 MB/s 6271234048 bytes (6.3 GB) copied, 191.075 s, 32.8 MB/s 6384636928 bytes (6.4 GB) copied, 195.988 s, 32.6 MB/s 6500127744 bytes (6.5 GB) copied, 200.907 s, 32.4 MB/s 6646800384 bytes (6.6 GB) copied, 205.778 s, 32.3 MB/s 6785296384 bytes (6.8 GB) copied, 210.863 s, 32.2 MB/s 6935811072 bytes (6.9 GB) copied, 215.973 s, 32.1 MB/s 7099803648 bytes (7.1 GB) copied, 220.938 s, 32.1 MB/s 7262974976 bytes (7.3 GB) copied, 225.935 s, 32.1 MB/s 7426348032 bytes (7.4 GB) copied, 231.052 s, 32.1 MB/s 7568887808 bytes (7.6 GB) copied, 236.058 s, 32.1 MB/s 7729822720 bytes (7.7 GB) copied, 240.99 s, 32.1 MB/s 7888008192 bytes (7.9 GB) copied, 246.025 s, 32.1 MB/s 8071897088 bytes (8.1 GB) copied, 251.018 s, 32.2 MB/s 8240689152 bytes (8.2 GB) copied, 256.032 s, 32.2 MB/s 8391177216 bytes (8.4 GB) copied, 261.789 s, 32.1 MB/s 8529931264 bytes (8.5 GB) copied, 266.152 s, 32.0 MB/s 8692167680 bytes (8.7 GB) copied, 271.163 s, 32.1 MB/s 8828957696 bytes (8.8 GB) copied, 276.153 s, 32.0 MB/s 9006211072 bytes (9.0 GB) copied, 281.133 s, 32.0 MB/s 9176976384 bytes (9.2 GB) copied, 286.11 s, 32.1 MB/s 9349107712 bytes (9.3 GB) copied, 291.283 s, 32.1 MB/s 9512068096 bytes (9.5 GB) copied, 296.149 s, 32.1 MB/s 9686129664 bytes (9.7 GB) copied, 301.171 s, 32.2 MB/s 9807513600 bytes (9.8 GB) copied, 306.388 s, 32.0 MB/s 9964740608 bytes (10 GB) copied, 311.281 s, 32.0 MB/s 10107637760 bytes (10 GB) copied, 316.382 s, 31.9 MB/s 10240000000 bytes (10 GB) copied, 319.844 s, 32.0 MB/s write complete, syncing reading from: /mnt/disk3/Movies/speedtest3 10240000000 bytes (10 GB) copied, 125.35 s, 81.7 MB/s removing: /mnt/disk3/Movies/speedtest3 removed `/mnt/disk3/Movies/speedtest3' Quote Link to comment
Joe L. Posted January 20, 2011 Share Posted January 20, 2011 Nice utility. You can really see how the initial portion of the file copied to RAM disk cache buffers skews the overall speed for the disk. No meaningful compare between systems can therefore be made unless they have the same amount of RAM. (not if only using the final overall speed reported by "dd") Joe L. Quote Link to comment
WeeboTech Posted January 20, 2011 Share Posted January 20, 2011 Geeze those initial numbers are good. That is what I was going for. I guess I did not reach it. Usually when I rip dvd's I have an ITX machine rip directly with anydvd to to a drive on the unRAID server. On the other side, when I download stuff from the internet, I download it directly to the server. So the initial burst is important to me. I know my machine eventually slows down to 45MB/s on the long haul. As Joe mentions the final speed at the end is what matters to most people. How fast and how long it takes to move files. Quote Link to comment
BRiT Posted January 20, 2011 Share Posted January 20, 2011 I think the initial numbers are only that good because most of my system's ram was available for buffering along with the newer kernels having improved flushing system over the native unRAID version (2.6.32.x). I think that was introduced in 2.6.33 or 2.6.34 series. Here's what the system looks like with no active virtuals, handbrake jobs or downloads (nttp/p2p). Linux reaver 2.6.36.3-unraid #2 SMP Sun Jan 9 22:14:53 EST 2011 x86_64 Intel(R) Core(TM) i3 CPU 530 @ 2.93GHz GenuineIntel GNU/Linux procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu---- r b swpd free buff cache si so bi bo in cs us sy id wa 0 0 0 2462396 385820 498744 0 0 2072 735 4 4 1 1 94 4 total used free shared buffers cached Mem: 3787192 1324796 2462396 0 385820 498744 -/+ buffers/cache: 440232 3346960 Swap: 8000364 0 8000364 13:01:05 up 8 days, 17:20, 2 users, load average: 0.04, 0.04, 0.05 Quote Link to comment
prostuff1 Posted February 19, 2011 Share Posted February 19, 2011 Thanks for this little tool. I am running it on some of the new Hitachi "coolspin" drives. Check out the 2TB drives thread in the Hard Drives and Controllers forum. Quote Link to comment
Fuggin Posted December 3, 2013 Share Posted December 3, 2013 When doing the diskspeed.sh test o each drive, should the parity drive be disabled? Initial test. I am intrigued by the read speeds root@Tower:/boot# diskspeed.sh /mnt/disk2/testspeed writing 10240000000 bytes to: /mnt/disk2/testspeed 160149+0 records in 160149+0 records out 163992576 bytes (164 MB) copied, 5.29052 s, 31.0 MB/s 346865+0 records in 346865+0 records out 355189760 bytes (355 MB) copied, 10.1569 s, 35.0 MB/s 517324+0 records in 517324+0 records out 529739776 bytes (530 MB) copied, 15.0922 s, 35.1 MB/s 556193+0 records in 556193+0 records out 569541632 bytes (570 MB) copied, 20.7465 s, 27.5 MB/s 676015+0 records in 676015+0 records out 692239360 bytes (692 MB) copied, 25.1497 s, 27.5 MB/s 734041+0 records in 734041+0 records out 751657984 bytes (752 MB) copied, 31.958 s, 23.5 MB/s 811477+0 records in 811477+0 records out 830952448 bytes (831 MB) copied, 35.2513 s, 23.6 MB/s 912749+0 records in 912749+0 records out 934654976 bytes (935 MB) copied, 40.4368 s, 23.1 MB/s 1146021+0 records in 1146021+0 records out 1173525504 bytes (1.2 GB) copied, 45.24 s, 25.9 MB/s 1293685+0 records in 1293685+0 records out 1324733440 bytes (1.3 GB) copied, 50.2709 s, 26.4 MB/s 1459757+0 records in 1459757+0 records out 1494791168 bytes (1.5 GB) copied, 55.3769 s, 27.0 MB/s 1657499+0 records in 1657499+0 records out 1697278976 bytes (1.7 GB) copied, 60.3168 s, 28.1 MB/s 1831967+0 records in 1831967+0 records out 1875934208 bytes (1.9 GB) copied, 65.3439 s, 28.7 MB/s 2027513+0 records in 2027513+0 records out 2076173312 bytes (2.1 GB) copied, 70.3769 s, 29.5 MB/s 2226529+0 records in 2226529+0 records out 2279965696 bytes (2.3 GB) copied, 75.3928 s, 30.2 MB/s 2446305+0 records in 2446305+0 records out 2505016320 bytes (2.5 GB) copied, 80.4269 s, 31.1 MB/s 2581269+0 records in 2581269+0 records out 2643219456 bytes (2.6 GB) copied, 85.4469 s, 30.9 MB/s 2785343+0 records in 2785343+0 records out 2852191232 bytes (2.9 GB) copied, 90.4802 s, 31.5 MB/s 2961358+0 records in 2961358+0 records out 3032430592 bytes (3.0 GB) copied, 95.5028 s, 31.8 MB/s 3141389+0 records in 3141389+0 records out 3216782336 bytes (3.2 GB) copied, 100.537 s, 32.0 MB/s 3337461+0 records in 3337461+0 records out 3417560064 bytes (3.4 GB) copied, 105.577 s, 32.4 MB/s 3532974+0 records in 3532974+0 records out 3617765376 bytes (3.6 GB) copied, 111.255 s, 32.5 MB/s 3707097+0 records in 3707097+0 records out 3796067328 bytes (3.8 GB) copied, 115.62 s, 32.8 MB/s 3907049+0 records in 3907049+0 records out 4000818176 bytes (4.0 GB) copied, 120.657 s, 33.2 MB/s 4073070+0 records in 4073070+0 records out 4170823680 bytes (4.2 GB) copied, 125.67 s, 33.2 MB/s 4277937+0 records in 4277937+0 records out 4380607488 bytes (4.4 GB) copied, 130.697 s, 33.5 MB/s 4436448+0 records in 4436448+0 records out 4542922752 bytes (4.5 GB) copied, 135.729 s, 33.5 MB/s 4616958+0 records in 4616958+0 records out 4727764992 bytes (4.7 GB) copied, 140.745 s, 33.6 MB/s 4800157+0 records in 4800157+0 records out 4915360768 bytes (4.9 GB) copied, 145.772 s, 33.7 MB/s 4979413+0 records in 4979413+0 records out 5098918912 bytes (5.1 GB) copied, 150.797 s, 33.8 MB/s 5157528+0 records in 5157528+0 records out 5281308672 bytes (5.3 GB) copied, 155.82 s, 33.9 MB/s 5356824+0 records in 5356824+0 records out 5485387776 bytes (5.5 GB) copied, 160.838 s, 34.1 MB/s 5564395+0 records in 5564395+0 records out 5697940480 bytes (5.7 GB) copied, 165.868 s, 34.4 MB/s 5756749+0 records in 5756749+0 records out 5894910976 bytes (5.9 GB) copied, 170.908 s, 34.5 MB/s 5940461+0 records in 5940461+0 records out 6083032064 bytes (6.1 GB) copied, 175.948 s, 34.6 MB/s 6110632+0 records in 6110632+0 records out 6257287168 bytes (6.3 GB) copied, 180.981 s, 34.6 MB/s 6286700+0 records in 6286700+0 records out 6437580800 bytes (6.4 GB) copied, 186.002 s, 34.6 MB/s 6491665+0 records in 6491665+0 records out 6647464960 bytes (6.6 GB) copied, 191.047 s, 34.8 MB/s 6666874+0 records in 6666873+0 records out 6826877952 bytes (6.8 GB) copied, 196.06 s, 34.8 MB/s 6876237+0 records in 6876237+0 records out 7041266688 bytes (7.0 GB) copied, 201.091 s, 35.0 MB/s 7029783+0 records in 7029782+0 records out 7198496768 bytes (7.2 GB) copied, 206.146 s, 34.9 MB/s 7223421+0 records in 7223421+0 records out 7396783104 bytes (7.4 GB) copied, 211.308 s, 35.0 MB/s 7393485+0 records in 7393485+0 records out 7570928640 bytes (7.6 GB) copied, 216.203 s, 35.0 MB/s 7561162+0 records in 7561162+0 records out 7742629888 bytes (7.7 GB) copied, 221.232 s, 35.0 MB/s 7742673+0 records in 7742673+0 records out 7928497152 bytes (7.9 GB) copied, 226.297 s, 35.0 MB/s 7928261+0 records in 7928261+0 records out 8118539264 bytes (8.1 GB) copied, 231.411 s, 35.1 MB/s 8112641+0 records in 8112641+0 records out 8307344384 bytes (8.3 GB) copied, 236.347 s, 35.1 MB/s 8296677+0 records in 8296677+0 records out 8495797248 bytes (8.5 GB) copied, 241.387 s, 35.2 MB/s 8466439+0 records in 8466439+0 records out 8669633536 bytes (8.7 GB) copied, 246.412 s, 35.2 MB/s 8633745+0 records in 8633745+0 records out 8840954880 bytes (8.8 GB) copied, 251.437 s, 35.2 MB/s 8792601+0 records in 8792601+0 records out 9003623424 bytes (9.0 GB) copied, 256.459 s, 35.1 MB/s 8965157+0 records in 8965157+0 records out 9180320768 bytes (9.2 GB) copied, 261.491 s, 35.1 MB/s 9138653+0 records in 9138653+0 records out 9357980672 bytes (9.4 GB) copied, 266.521 s, 35.1 MB/s 9263738+0 records in 9263738+0 records out 9486067712 bytes (9.5 GB) copied, 271.547 s, 34.9 MB/s 9407825+0 records in 9407825+0 records out 9633612800 bytes (9.6 GB) copied, 276.564 s, 34.8 MB/s 9548933+0 records in 9548933+0 records out 9778107392 bytes (9.8 GB) copied, 281.627 s, 34.7 MB/s 9688509+0 records in 9688509+0 records out 9921033216 bytes (9.9 GB) copied, 286.619 s, 34.6 MB/s 9810853+0 records in 9810853+0 records out 10046313472 bytes (10 GB) copied, 291.667 s, 34.4 MB/s 9924441+0 records in 9924441+0 records out 10162627584 bytes (10 GB) copied, 296.747 s, 34.2 MB/s 10000000+0 records in 10000000+0 records out 10240000000 bytes (10 GB) copied, 299.968 s, 34.1 MB/s write complete, syncing reading from: /mnt/disk2/testspeed 10000000+0 records in 10000000+0 records out 10240000000 bytes (10 GB) copied, 8.7952 s, 1.2 GB/s removing: /mnt/disk2/testspeed removed `/mnt/disk2/testspeed' root@Tower:/boot# Quote Link to comment
garycase Posted December 3, 2013 Share Posted December 3, 2013 Well this is weird. My WD Caviar Black drive is apparently slower than my WD20EARS drives. THAT is interesting. Any reason why?? Sure. Note that "slower" depends on what you're measuring. A WD Black will definitely have better seek times, so if you were doing a lot of IOPS with small blocks it would outperform the EARS units. But what you're measuring here is sustained data transfer rates for fairly large transfers => so the timing depends on two things: (1) the rotational speed (the Black clearly "wins"); and (2) the areal density (the EARS win). Without specific model #'s, I can't say for sure, but I'd guess your Black is a 3-platter unit (333GB/platter) whereas your EARS are 500GB/platter units .. so the higher density of the EARS results in faster transfer rates despite the faster rotation of the Black. Quote Link to comment
garycase Posted December 3, 2013 Share Posted December 3, 2013 Note that if your Black was one of the new WD1003FZEX drives it'd "smoke" your EARS drives (This is a 1TB/platter drive) Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.