Re: [opensuse] Hard disc questions Slight OT

On Sun, Nov 23, 2008 at 6:52 AM, peter nikolic
<p.nikolic1@xxxxxxxxxxxxxx> wrote:
Hi .

I have been noticing this system is somewhat slow on the drive i use for music
and video storage

the drive is an SAMSUNG HD403LJ 400Gb Sata and it seems very slow

7-of-9:/data5 # time dd if=/dev/zero of=test bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 12.9316 s, 83.0 MB/s

real 0m13.004s
user 0m0.008s
sys 0m2.956s

now that compared with the main system drive which is an ST3200826A

7-of-9:/ # time dd if=/dev/zero of=test bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 9.57748 s, 112 MB/s

real 0m9.581s
user 0m0.004s
sys 0m4.300s

seems very slow for a drive that is a lot newer

anyone seeing this or got any ideas

System is Asrock Mboard KS8Xxxxx 64bit AMD Athlon64 CPU
2Gb ram Nvidia FX5500 @256 Mb
Opensuse 10.3 x86_64 KDE 3.5.7 "release 72.9"
Kernel 2.6.22.19-0.1-default

Not the best or fastest on MoBo's but this from the Sata drive is poor

Cheerrs Pete .

First 83 MB/sec is about 5GB a minute. That is a very respectable
speed, but I can see why you would like to get 112 MB/s, which would
be about 6.5 GB a minute.

Hard to say what is going on, but a few issues with your benchmark.

1) /dev/zero in 11.0 (iirc) is buggy and causes performance
limitations. Should not be used for bench mark purposes, especially
at speed as high as you are getting.

2) Inner radius of the physical drive will transmit at a different
speed than outer radius. iirc, the outer radius is faster because the
RPMs are the same, but they pack more data onto the outer tracks than
they do on the inner tracks so you get more data per revolution.

Since you are writing to a file, you don't have anyway to say what
part of the drive you are writing to. And yes, the variation can be
pretty big. Try performing your process to an entire drive (dd
if=/dev/zero of=/dev/sdb ...) And then use iostat -d 5 to watch the
transfer speed. Assuming /dev/zero is not limiting you, you should
see the transfer rate slowly decrease as you continue writing data.

3) You don't say if it is the same filesystem / PCI-bus / cabling /
disk controller / etc. At 100MB/sec you are really pushing all of the
components hard.

4) A 1 GB test is way to small to mean much. Lots of caches involved
that can interfere with that small of a test. ie. If your disk has a
256MB write cache, it will report done when you're really only 75%
done. Kernel cache can be even bigger. Add them together and is is
very conceivable that well over half of your 15 second test is just
pushing data to cache, and it is taking another 15 seconds to flush
those caches. Thus in reality you are only getting half the speed you
think you are.

Good luck tracking in down, but I suspect it is much more of a complex
issue to even benchmark than you suspect.

Greg
--
Greg Freemyer
Litigation Triage Solutions Specialist
http://www.linkedin.com/in/gregfreemyer
First 99 Days Litigation White Paper -
http://www.norcrossgroup.com/forms/whitepapers/99%20Days%20whitepaper.pdf

The Norcross Group
The Intersection of Evidence & Technology
http://www.norcrossgroup.com
--
To unsubscribe, e-mail: opensuse+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: opensuse+help@xxxxxxxxxxxx

Relevant Pages

  • RE: Disk vs Tape scenario
    ... With the storage available for me to test it is hard to eliminate cache ... Theoretically a volume using RAID-10 on 8x10K RPM disks can handle ... A write SSCH rate of 475/sec would push the RAID-10 parity group to 950 ... I just tested with two IEBDG jobs writing 20 million 80 byte records ...
    (bit.listserv.ibm-main)
  • Re: Cache questions
    ... > sets and the MMX and XMM registers where possible. ... > I'm writing a routine which will need to read from a source bitmap ... especially when it causes cache misses. ... > some stack space, and I could write one of my local variables with, ...
    (comp.lang.asm.x86)
  • Re: need fastest way to write 2gig array to disk file
    ... Eric Taylor wrote: ... > writes seem to go directly to the cache and so a 2 gig output ... > On another similar system, with scsi disks, once the program ... (when writing to the cache, ...
    (comp.os.linux.development.apps)
  • Cache questions
    ... sets and the MMX and XMM registers where possible. ... I'm writing a routine which will need to read from a source bitmap ... But actually my knowledge of how the L1/L2 cache works is a little ... my local variables to memory, what's the best way to do that? ...
    (comp.lang.asm.x86)