View previous topic :: View next topic |
Author |
Message |
rbr28 Tux's lil' helper
Joined: 09 Feb 2004 Posts: 126
|
Posted: Thu Feb 12, 2004 8:35 pm Post subject: SATA - hdparm - performance |
|
|
I've seen a lot of comments about drive performance and a good bit about SATA, but I have yet to see a good summary. If I missed the thread someone please point me there.
I'm using the 2.6.1 kernel right now, from gentoo-dev-sources. I have two 80GB Seagate SATA drives and I get an ourtageous number of around 1400MB/sec on hdparm -t and 55MB/sec on hdparm -T. I'm lazy and not doublechecking, so I might have those reversed. Anyway, from what I have seen those are good numbers but I'm really not sure since I haven't tested many other SATA systems yet. I have tested other SCSI workstations and other ATA workstations and I have never seen above 38MB/s for the lower number, but I would just be interested in how others are making out with the 2.6 kernel and SATA, and if there are any optimizations that have worked well.
Additionally, I disabled the option for dma by default when I configured the kernel, to see if that was related to an early problem I was having. I no longer have the problem but I changed several things at once and I was wondering if disabling dma by default was necessary and/or the preferred setting for SATA drives and the 2.6 kernel. |
|
Back to top |
|
|
Dr_b_ n00b
Joined: 18 Jan 2004 Posts: 33
|
Posted: Sat Feb 14, 2004 4:08 pm Post subject: |
|
|
What liveCD did you use to get your system to recognize your SATA Drive?
What is the drive showing up as? /dev/hd? One one of the liveCD's its showing up as a SCSI device. The latest experimentals wont boot my system they seem to hang at some point. |
|
Back to top |
|
|
arska n00b
Joined: 23 Aug 2003 Posts: 10 Location: San José, Costa Rica
|
Posted: Sat Feb 14, 2004 6:25 pm Post subject: |
|
|
My config is following:
- DMA set in 2.6.3-rc2 kernel.
- Linux assigns /dev/sda for the HDD
- Same hdparm results |
|
Back to top |
|
|
rbr28 Tux's lil' helper
Joined: 09 Feb 2004 Posts: 126
|
Posted: Sat Feb 14, 2004 8:13 pm Post subject: sata setup -Asus P4C800-E Deluxe |
|
|
I'm going to post a more complete mini howto to help other people but for now, here are some pointers. I use the Mandrake Move cd to build my system. The latest one recognized my SATA drives without any problem. My setup is an Asus P4C800-E Deluxe, with two SATA drives on the Intel ICH5R controller, not in a raid setup. The bios is set to run the drives in native mode. I have two CD-roms connected and I have the promise raid controller disabled in the bios.
When i boot with the MandrakeMove cd, my SATA drives are mounted as hde and hdf. One note, when I used the Mandrake cd and followed the instructions for Knoppix, I had problems with the install. Follow the instructions for using the LiveCD for the most part. Make sure you mount /proc and /dev the way they tell you to do it for the LiveCD, not the way they say to do it for the Knoppix based install.
Anyway you can find info all over the forum about what you need to compile into your kernel, depending on which one you are using. I had no problem with anthing from 2.4.22 on and I finally settled on the gentoo-dev-sources when they were using the 2.6.1 kernel. I also compiled the mm-sources when they were using the 2.6.3 kernel, but I had some minor issues that I didn't want to bother working through, since I had the 2.6.1 kernel working flawlessly. The drives worked fine with all the kernels I tried though. I measured performance and I had no significant difference between the 2.6.1 kernel and the gentoo-dev-sources, and the mm-sources with a slightly newer kernel. My final performance stats were around 55MB/s and 2000MB/s for the two different hdparm -tT /dev/sda numbers. I ran piozone and came up with very similar numbers.
Anyway follow everyone elses recommendations about compiling scsi emulation, scsi-generic, ata, serial ata, and so on into your kernel, don't compile these as modules. Make sure you compile the specific drivers for your setup too. For mine, it was the Piix drivers, or something like that, and the intel driver under SATA (which is under scsi low level drivers).
One of the problems I had was that Mandrake Move recognizes the drives as hde and hdd as I mentioned. The kernels (at least 2.6.x kernels) will recognize your drives as /dev/sdxx . Make sure you setup your fstab and your grub.conf to point to the appropriate partitions such as /dev/sda1, rather than /dev/hde1 . If your grub.conf is not configured correctly you will never get it to work. Remember too, if you mess it up, you can get a command shell from grub to edit the lines in grub.conf before you boot. If your fstab is incorrectly setup though, you will still have problems and you will need to do something like booting with whatever cd based distro you used, and then edit the file from there.
What I've seen is that most people are getting the right drivers compiled in, because there is lots of info on that all over the forum. Most people are making the mistakes in setting up grub.conf or fstab. Also, don't use genkernel. It just introduces another point of failure and makes troubleshooting more difficult.
Finally if you get all that working, you'll probably have trouble with the cdrom drives and/or cd-writers if you have those too. I had to put hdc=ide-scsi and hdd=ide-scsi on my kernel line in grub for both my cd-rom drives to work correctly. I also had to edit my fstab to mount /dev/cdroms/cdrom0 and /dev/cdroms/cdrom1 respectively. This allowed my burners to work both with burning software like K3B and with programs like XMMS.
I wish kernel developers would get the whole ide/scsi thing worked out. It's very confusing using different kernels and trying to figure whether something needs to be loaded as a module or not, whether to use scsi, scsi emulation , ata, sata, etc...how things are referenced by what, how they are mounted, etc. Seems to change all the time with different kernels and with different hardware. It gets easier the more you have messed with it, but I can imagine how frustrating it is for first timers.
Anyway, dont give up. It took me weeks, but I got the Gentoo-dev-sources with the 2.6.1 kernel running flawlessly on my board. It's probably the best performing machine I have ever used, and definitely the best performing linux machine I have ever used (not including servers of course). I have every single thing working flawlessly, no errors whatsoever on bootup. Gigabit ethernet was no problem, sound works great, video is awesome (Nvidia Geforce), drive performance is great (including the burners), etc. It's taken a lot of work, but it's been worth it. |
|
Back to top |
|
|
Dr_b_ n00b
Joined: 18 Jan 2004 Posts: 33
|
Posted: Sun Feb 15, 2004 2:20 am Post subject: |
|
|
Thats an awesome post, thanks for the reply, I have by no means given up.
Look forward to your SATA howto.
The gentoo-2004.0-x86-20040121.iso liveCD experimental image actually works too. I also have the Asus P4C800-E deluxe mainboard and WD raptors.
I agree with your this whole scsi-ide thing is a bit confusing as Im not yet very familiar with how to set it up, for instance i couldn't figure out why an IDE drive would need to be recognized as a SCSI device. |
|
Back to top |
|
|
DarrenM l33t
Joined: 25 Apr 2002 Posts: 653 Location: Sydney, Australia
|
Posted: Wed Feb 18, 2004 2:41 am Post subject: |
|
|
I just got a couple of SATA Raptors and wasn't too impressed at my test hdparm figures. I'm hoping I've missed something somewhere.
ATA-66 drive -t 34MB/s
ATA-100 drive -t 38MB/s
SATA Raptor -t 40MB/s
all get 900MB/s with -T |
|
Back to top |
|
|
nmcsween Guru
Joined: 12 Nov 2003 Posts: 381
|
Posted: Wed Feb 18, 2004 9:28 am Post subject: |
|
|
maybe it's just me but all those numbers seem a little low. I get about 980mb and 43mb with hdparm on one sata disk. |
|
Back to top |
|
|
serotonin n00b
Joined: 24 Dec 2003 Posts: 50
|
Posted: Thu Feb 19, 2004 6:14 am Post subject: |
|
|
im a noob when it comes to tweaking hard drives, do i have this thing running well ?
here's my result with the onboard via sata controller on an xp 2700+, 512 ddr3200
segate 160 gb baracuda
hdparm -i /dev/hde
/dev/hde:
Model=ST3160023AS, FwRev=3.18, SerialNo=3JS2E9YT
Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs RotSpdTol>.5% }
RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4
BuffType=unknown, BuffSize=8192kB, MaxMultSect=16, MultSect=16
CurCHS=65535/1/63, CurSects=4128705, LBA=yes, LBAsects=268435455
IORDY=on/off, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120}
PIO modes: pio0 pio1 pio2 pio3 pio4
DMA modes: mdma0 mdma1 mdma2
UDMA modes: udma0 udma1 udma2
AdvancedPM=no WriteCache=enabled
Drive conforms to: ATA/ATAPI-6 T13 1410D revision 2:
* signifies the current active mode
hdparm -T /dev/hde
/dev/hde:
Timing buffer-cache reads: 980 MB in 2.00 seconds = 488.85 MB/sec
hdparm -t /dev/hde
/dev/hde:
Timing buffered disk reads: 152 MB in 3.01 seconds = 50.56 MB/sec
last,
hdparm -I /dev/hde
/dev/hde:
ATA device, with non-removable media
Model Number: ST3160023AS
Serial Number: 3JS2E9YT
Firmware Revision: 3.18
Standards:
Used: ATA/ATAPI-6 T13 1410D revision 2
Supported: 6 5 4 3
Configuration:
Logical max current
cylinders 16383 65535
heads 16 1
sectors/track 63 63
--
CHS current addressable sectors: 4128705
LBA user addressable sectors: 268435455
LBA48 user addressable sectors: 312581808
device size with M = 1024*1024: 152627 MBytes
device size with M = 1000*1000: 160041 MBytes (160 GB)
Capabilities:
LBA, IORDY(can be disabled)
bytes avail on r/w long: 4 Queue depth: 1
Standby timer values: spec'd by Standard
R/W multiple sector transfer: Max = 16 Current = 16
Recommended acoustic management value: 254, current value: 0
DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 udma5 *udma6
Cycle time: min=120ns recommended=120ns
PIO: pio0 pio1 pio2 pio3 pio4
Cycle time: no flow control=240ns IORDY flow control=120ns
Commands/features:
Enabled Supported:
* READ BUFFER cmd
* WRITE BUFFER cmd
* Host Protected Area feature set
* Look-ahead
* Write cache
* Power Management feature set
Security Mode feature set
* SMART feature set
* FLUSH CACHE EXT command
* Mandatory FLUSH CACHE command
* Device Configuration Overlay feature set
* 48-bit Address feature set
SET MAX security extension
* DOWNLOAD MICROCODE cmd
* SMART self-test
* SMART error logging
Security:
supported
not enabled
not locked
not frozen
not expired: security count
not supported: enhanced erase
Checksum: correct
i have yet to try with the scsi drivers. any good ? _________________ http://www.tankprofiler.com
Last edited by serotonin on Thu Feb 19, 2004 6:20 am; edited 2 times in total |
|
Back to top |
|
|
Moled l33t
Joined: 09 Jul 2003 Posts: 635
|
Posted: Thu Feb 19, 2004 6:16 am Post subject: |
|
|
I get 2000+ / 50-55 |
|
Back to top |
|
|
Crg Guru
Joined: 29 May 2002 Posts: 345 Location: London
|
Posted: Thu Feb 19, 2004 12:13 pm Post subject: |
|
|
DarrenM wrote: | I just got a couple of SATA Raptors and wasn't too impressed at my test hdparm figures. I'm hoping I've missed something somewhere.
ATA-66 drive -t 34MB/s
ATA-100 drive -t 38MB/s
SATA Raptor -t 40MB/s
all get 900MB/s with -T |
Not surprising that they all get 900MB/s with -T as it is testing the speed of reading directly from linux's buffer cache - ie it isn't testing the drives.
The SATA drivers for linux are quite new so there might be more performance tweaking that can be done, but I wouldn't expect it to be much faster than ATA-133, and the performance really comes done to the disk drive itself.
For example a Barracuda 7200RPM 120GB ATA-100 has a sustained avg transfer rate of >58Mbytes/sec, where as a Barracudua 7200RPM Serial ATA 120GB only has an avg sustained rate of >44MBytes/sec. |
|
Back to top |
|
|
TheJackal n00b
Joined: 29 Jul 2003 Posts: 73 Location: In front of the keyboard :)
|
Posted: Mon May 24, 2004 10:03 pm Post subject: |
|
|
Just thout I'd add my 50 cents worth :
Code: | /dev/sda:
Timing buffer-cache reads: 1180 MB in 2.01 seconds = 588.32 MB/sec
Timing buffered disk reads: 166 MB in 3.00 seconds = 55.29 MB/sec |
Seagate 7200.7 80GB (ST380013AS) 8MB cache SATA and Kernel 2.6.5 (gentoo-dev-sources)
With my old Seagate Barracuda 80GB ATAIV (ATA100, 7200rpm, 2MB cache) - same kernel, I used to get :
Code: | /dev/hda:
Timing buffer-cache reads: 728 MB in 2.01 seconds = 362.60 MB/sec
Timing buffered disk reads: 120 MB in 3.04 seconds = 39.43 MB/sec |
_________________ "Expect nothing. That way you will be ready for anything - Virginia Viadura"
Black Mesa Research Facility |
|
Back to top |
|
|
torklingberg Tux's lil' helper
Joined: 30 May 2004 Posts: 86
|
Posted: Thu Jun 24, 2004 3:34 pm Post subject: |
|
|
A SATA HOWTO would be really nice. It is i really a mess. |
|
Back to top |
|
|
eNTi Veteran
Joined: 20 Oct 2002 Posts: 1011 Location: Salzburg, Austria
|
Posted: Thu Jun 24, 2004 8:08 pm Post subject: |
|
|
Code: | /dev/sda:
Timing buffer-cache reads: 1012 MB in 2.00 seconds = 505.07 MB/sec
Timing buffered disk reads: 156 MB in 3.01 seconds = 51.80 MB/sec
|
is there a way to optimize a sata drive?
i got a Segate 160GB 7200rpm SATA UDMA6 (on PCI Promise Controller SATA150 TX2plus). _________________ If you fall off a cliff, you might as well try to fly. After all, you got nothing to lose.
-- John Sheridan - Babylon 5, Season 4
Last edited by eNTi on Sun May 07, 2006 3:36 pm; edited 1 time in total |
|
Back to top |
|
|
serendipity n00b
Joined: 05 Jun 2004 Posts: 69
|
Posted: Fri Jun 25, 2004 10:37 am Post subject: |
|
|
I have a working ebuild to emerge patched 2.4.26 sources that handle SATA and SATA RAID 0 on the ICH5-R. I'm still working on the genkernel and grub mods, which is why nothing is posted yet, but with these three ebuilds, it should be possible to
emerge i875p-iswraid-sources
emerge grub-iswraid
grub-install /dev/ataraid/d0
emerge genkernel-iswraid
genkernel
... modify grub.conf to load the new kernel and initrd
and reboot.
kernel 2.4.26 with
- device mapper patches
- libata patch
- iswraid patch
- i2c and lm_sensors 2.8.7
- video4linux patches
The kernel source is the easy part, although the config (modules vs compiled in) can be tricky. The difficult part is genkernel, which needs to be hacked to create an appropriate initrd (problems with device files, module probing, errors in devfsd.conf). I'm working on a genkernel-iswraid ebuild.
Also, if you boot off a raid array, grub needs to be patched. I'm working on an ebuild for a grub-iswraid.
If you want the kernel source ebuild, and you are technically oriented (I'll send you the genkernel diffs), let me know and I'll send you the ebuild. Don't expect much support. If you need support, then wait until I 've finished the ebuilds so that you can just emerge everything blindly.
Oh, and I also posted a nasty hack to hdparm for it to benchmark ataraid drives. Look on this forum for "hdparam ataraid patch". Im getting 95MB/s off my two raptors in raid 0 in the ich5-r controller. |
|
Back to top |
|
|
BlinkEye Veteran
Joined: 21 Oct 2003 Posts: 1046 Location: Gentoo Forums
|
Posted: Fri Jun 25, 2004 11:49 am Post subject: |
|
|
it's about time someone started a thread about it. i almost started one but i don't have enough time right now. so just in short (some of us already started a dispute here: https://forums.gentoo.org/viewtopic.php?t=8813&postdays=0&postorder=asc&start=75)
i don't think hdparm returns any accurate result at all about your hardware performance. as suggested in the link above try out IOZone to measure your performance and compare them with others. i collected 2 different IOZone files from other systems and users with a raid and SATA hd's and of course made some test myself - even though my drives seem to be much slower according to "hdparm -t -T blabla" i won most of the test in IOZone ( http://www.iozone.org/ - have a look at the above mentioned thread as i posted some command on how to use it). we still could compare our IOZone files, i still in the mood to compete! _________________ Easily backup up your system? klick
Get rid of SSH Brute Force Attempts / Script Kiddies klick |
|
Back to top |
|
|
Nephren-Ka n00b
Joined: 10 Jun 2002 Posts: 63 Location: Melbourne, Australia
|
Posted: Fri Jun 25, 2004 6:15 pm Post subject: |
|
|
I have two sata drives:
Seagate 160:
hdparm -Tt /dev/sda
Code: |
/dev/sda:
Timing buffer-cache reads: 3872 MB in 2.00 seconds = 1935.33 MB/sec
Timing buffered disk reads: 158 MB in 3.03 seconds = 52.19 MB/sec
|
****TheJackal, enti, and others: Note, this is the same drive you guys have, and your buffer-cache read results are WAY too slow, you have something set up improperly, I'd check that out....
*****
and Seagate 200:
Code: |
hdparm -Tt /dev/sdb
/dev/sdb:
Timing buffer-cache reads: 3840 MB in 2.00 seconds = 1920.29 MB/sec
Timing buffered disk reads: 188 MB in 3.03 seconds = 62.06 MB/sec
|
Setting them up was really quite simple (I have an ABIT IS7-G, Intel i865-PE chipset)...simply build the SATA support into the kernel (in the SCSI setup part of the kernel config), and make sure you have i865 and ICH5 support compiled in as well....and you're golden If anyone has a simliar setup and needs help, don't hesitate to ask [/b] |
|
Back to top |
|
|
BlinkEye Veteran
Joined: 21 Oct 2003 Posts: 1046 Location: Gentoo Forums
|
Posted: Fri Jun 25, 2004 6:38 pm Post subject: |
|
|
i repeat, i do have 3 seagate sata drives in a raid 5 too and do not get anywhere near your result. this isn't a setting problem but a testing issue. hdparm doesn't say if your drives are fast or not (altough it says if dma is enabled or not - which results in a faster or slower drive - but don't relay on the result). use IOZone, a simple, small software which really tests your drives and post your results here (maybe do the same BIG test as i did (takes about an hour or so - follow the link of the thread in the gentoo forum) so we could really compare and after that we're going to discuss settings! or do you really relay on a test which takes about 5 seconds to see if your drive is fast or not? _________________ Easily backup up your system? klick
Get rid of SSH Brute Force Attempts / Script Kiddies klick |
|
Back to top |
|
|
Corona688 Veteran
Joined: 10 Jan 2004 Posts: 1204
|
Posted: Fri Jun 25, 2004 6:52 pm Post subject: |
|
|
BlinkEye wrote: | i repeat, i do have 3 seagate sata drives in a raid 5 too and do not get anywhere near your result. this isn't a setting problem but a testing issue. hdparm doesn't say if your drives are fast or not (altough it says if dma is enabled or not - which results in a faster or slower drive - but don't relay on the result). use IOZone, a simple, small software which really tests your drives and post your results here (maybe do the same BIG test as i did (takes about an hour or so - follow the link of the thread in the gentoo forum) so we could really compare and after that we're going to discuss settings! or do you really relay on a test which takes about 5 seconds to see if your drive is fast or not? | There's a line at which benchmarks turn into pointless hard-drive torture. I'd rather stray on the side of possible inaccuracy than the side of unnecessary wear. _________________ Petition for Better 64-bit ATI Drivers - Sign Here
http://www.petitiononline.com/atipet/petition.html |
|
Back to top |
|
|
Nephren-Ka n00b
Joined: 10 Jun 2002 Posts: 63 Location: Melbourne, Australia
|
Posted: Fri Jun 25, 2004 7:02 pm Post subject: |
|
|
I'm not sure if you were talking to me or not, however...I am not here to discuss the merits (or lack thereof) of hdparm testing. All I was saying, is that comparing apples to apples (all of us with same drives (or close) and using the same tool) are getting very different results.
It is indeed a settings problem, because I was getting the same low performance numbers they were getting, when I did not have the proper options compiled into my kernel.
BlinkEye wrote: | i repeat, i do have 3 seagate sata drives in a raid 5 too and do not get anywhere near your result. this isn't a setting problem but a testing issue. hdparm doesn't say if your drives are fast or not (altough it says if dma is enabled or not - which results in a faster or slower drive - but don't relay on the result). use IOZone, a simple, small software which really tests your drives and post your results here (maybe do the same BIG test as i did (takes about an hour or so - follow the link of the thread in the gentoo forum) so we could really compare and after that we're going to discuss settings! or do you really relay on a test which takes about 5 seconds to see if your drive is fast or not? |
|
|
Back to top |
|
|
BlinkEye Veteran
Joined: 21 Oct 2003 Posts: 1046 Location: Gentoo Forums
|
Posted: Fri Jun 25, 2004 8:27 pm Post subject: |
|
|
Nephren-Ka wrote: | I'm not sure if you were talking to me or not, however...I am not here to discuss the merits (or lack thereof) of hdparm testing. All I was saying, is that comparing apples to apples (all of us with same drives (or close) and using the same tool) are getting very different results.
It is indeed a settings problem, because I was getting the same low performance numbers they were getting, when I did not have the proper options compiled into my kernel. |
well, that sounds interesting but i honestly don't know what should cause the low buffer-cache readings. but would you please tell us how you fixed that ... ? you mentioned some posts above that you had the same problems and fixed it by enabling some mb specific sata driver and the raid stuff - but i can't believe that this is the issue. how is someone able to not have enabled sata and raid support while running a sata raid? _________________ Easily backup up your system? klick
Get rid of SSH Brute Force Attempts / Script Kiddies klick |
|
Back to top |
|
|
BlinkEye Veteran
Joined: 21 Oct 2003 Posts: 1046 Location: Gentoo Forums
|
Posted: Fri Jun 25, 2004 8:32 pm Post subject: |
|
|
Corona688 wrote: | There's a line at which benchmarks turn into pointless hard-drive torture. I'd rather stray on the side of possible inaccuracy than the side of unnecessary wear. |
that's a point. i just wanted to suggest that before someone loses a restless night over a bad benchmark from hdparm he should use another benchmark. it annoyed me at last, i admit _________________ Easily backup up your system? klick
Get rid of SSH Brute Force Attempts / Script Kiddies klick |
|
Back to top |
|
|
Nephren-Ka n00b
Joined: 10 Jun 2002 Posts: 63 Location: Melbourne, Australia
|
Posted: Fri Jun 25, 2004 9:14 pm Post subject: |
|
|
I'm not running SATA RAID, I just have 2 standalone drives. However, the problem I had is that I didnt have the proper sata controller driver compiled in, so the kernel wasn't able to use DMA properly with the drives...I'll post my kernel .config if you guys want?
BlinkEye wrote: | Nephren-Ka wrote: | I'm not sure if you were talking to me or not, however...I am not here to discuss the merits (or lack thereof) of hdparm testing. All I was saying, is that comparing apples to apples (all of us with same drives (or close) and using the same tool) are getting very different results.
It is indeed a settings problem, because I was getting the same low performance numbers they were getting, when I did not have the proper options compiled into my kernel. |
well, that sounds interesting but i honestly don't know what should cause the low buffer-cache readings. but would you please tell us how you fixed that ... ? you mentioned some posts above that you had the same problems and fixed it by enabling some mb specific sata driver and the raid stuff - but i can't believe that this is the issue. how is someone able to not have enabled sata and raid support while running a sata raid? |
|
|
Back to top |
|
|
taskara Advocate
Joined: 10 Apr 2002 Posts: 3763 Location: Australia
|
Posted: Fri Jun 25, 2004 11:32 pm Post subject: |
|
|
bonnie is your friends.. swap hdparm for it _________________ Kororaa install method - have Gentoo up and running quickly and easily, fully automated with an installer! |
|
Back to top |
|
|
serendipity n00b
Joined: 05 Jun 2004 Posts: 69
|
Posted: Fri Jun 25, 2004 11:41 pm Post subject: |
|
|
I liked this iozone thing. Here are the results of 45 non-stop minutes of my two maxtor 120GBs being thrashed by iozone, max file size specified as 2GB. I'm not too sure how often I'd like to run it, because the disks really do take a beating....
http://perso.wanadoo.fr/ic/iozoneresults.html |
|
Back to top |
|
|
lbrtuk l33t
Joined: 08 May 2003 Posts: 910
|
Posted: Sat Jun 26, 2004 1:43 am Post subject: |
|
|
The problem with iozone is it's not filesystem independant. It works on top of the filesystem. Therefore someone using reiserfs will get totally different results to someone using ext3 and it will have little to do with the hardware. |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|