Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
my harddrive performance is sucking
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, 4, 5, 6, 7, 8, 9  Next  
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware
View previous topic :: View next topic  
Author Message
Malakin
Veteran
Veteran


Joined: 14 Apr 2002
Posts: 1692
Location: Victoria BC Canada

PostPosted: Wed Nov 12, 2003 9:09 pm    Post subject: Reply with quote

hdparm -T is timing the reading of the disk cache in system memory not out of the hard disks cache so the size of the hard disk cache is irrelevant. This is why people are getting around 1GB/s, if it was pulling it out of the hard disk cache then you'd be maxing out your interface speed of 100-150MB/s for most people.

Also I think the hard disk cache is invisible to the system anyways and therefore unreadable directly.
Back to top
View user's profile Send private message
MK
Tux's lil' helper
Tux's lil' helper


Joined: 27 Nov 2002
Posts: 97
Location: Bærum, Norway

PostPosted: Tue Nov 25, 2003 5:32 pm    Post subject: Reply with quote

Hey,

I got an old SCSI disk, and haven't optimized it at all. Any tips for enhancing performance?

/dev/sda:
Timing buffer-cache reads: 612 MB in 2.01 seconds = 305.13 MB/sec
Timing buffered disk reads: 118 MB in 3.00 seconds = 39.33 MB/sec
Back to top
View user's profile Send private message
gubbs
n00b
n00b


Joined: 28 Feb 2003
Posts: 35

PostPosted: Mon Dec 01, 2003 4:10 am    Post subject: hardware raid bechmarking? Reply with quote

Hit here.

I have a Compaq SMART 431 array with 4 Wide Ultra 3 SCSI disks in Raid 0 mode.

The device is:

bash-2.05b# cat /dev/ida/disc0/disc with the partitions showing as:
disc part1 part2 part3 part4

I noticed some people on other threads can benchmark /dev/sdX but I dont have that .. and when i try hdparm -tT /dev/ida/disc0/disc it rejects the device as 'not being allowed'.

Anyone help me out? Perhaps there's some devious way I can CAT 128 mb chunks and work out the speed by the time taken or some other bash wizardy?

G.
Back to top
View user's profile Send private message
christsong84
Veteran
Veteran


Joined: 06 Apr 2003
Posts: 1003
Location: GMT-8 (Spokane)

PostPosted: Sat Feb 07, 2004 8:05 am    Post subject: Reply with quote

dunno...was told this isn't very good for an SATA disk...I'm quite new to the whole optimizing my hard drive thing...what do I need to do?

Code:
[96p] psalms : /home/swells $ hdparm -i /dev/hda

/dev/hda:

 Model=ST380013AS, FwRev=3.18, SerialNo=3JV8GWG8
 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=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=156301488
 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

[96p] psalms : /home/swells $ hdparm -Tt /dev/hda

/dev/hda:
 Timing buffer-cache reads:   2160 MB in  2.00 seconds = 1078.01 MB/sec
 Timing buffered disk reads:   28 MB in  3.68 seconds =   7.61 MB/sec

_________________
while(true) {self.input(sugar);} :twisted:
Back to top
View user's profile Send private message
der_maddin
Apprentice
Apprentice


Joined: 22 Feb 2004
Posts: 209
Location: somewhere between my systems ..

PostPosted: Thu Feb 26, 2004 12:25 am    Post subject: Reply with quote

these are my stats :

Code:


# hdparm -vtT /dev/hda

dev/hda:
 multcount    = 16 (on)
 IO_support   =  3 (32-bit w/sync)
 unmaskirq    =  1 (on)
 using_dma    =  1 (on)
 keepsettings =  1 (on)
 readonly     =  0 (off)
 readahead    = 256 (on)
 geometry     = 19457/255/63, sectors = 312581808, start = 0
 Timing buffer-cache reads:   996 MB in  2.00 seconds = 497.58 MB/sec
 Timing buffered disk reads:  160 MB in  3.02 seconds =  52.94 MB/sec



is this really good ?
as you can see, i own a samsung sp1614n
( 160gb, 7200rpm, 8M, u133 )
but why is the udma6 mode not available ??
i have allready set the correct kernel option for my chipset ??

i am confused.
_________________
neoCortex maddin # uname -a
Linux neoCortex 2.6.14-gentoo-r5 #2 SMP Fri Jan 13 04:40:02 CET 2006 x86_64 AMD Athlon(tm) 64 Processor 3400+ AuthenticAMD GNU/Linux
Back to top
View user's profile Send private message
woody77
n00b
n00b


Joined: 06 Feb 2004
Posts: 69
Location: SC Mtns, CA, USA

PostPosted: Fri Feb 27, 2004 4:17 am    Post subject: Reply with quote

Box: IBM T30 Laptop, 1.8 Ghz P4m, 512MB
HD: Toshiba MK4019GAX, 40GB, 16MB Cache, WriteCache=Enabled

udma 100

-vtT

multcount = 16
io = 32bit
unmaskirq = on
using_dma = on
keepsettings = on
readadhed = 15

cache-reads = 731MB/sec
disk-reads = 26MB/sec

And that's all she wrote. That's the best that I see. I've played with a few other options, but that's about all I get. I discovered it was running in 16-bit mode, and that helped it get to where it was, along with upping the read-ahead to 15 from 8, but since then, 25/26 seems to be the limit. Not bad, for a laptop, I guess, but I really wanted to see more.
Back to top
View user's profile Send private message
wolfpakk
n00b
n00b


Joined: 19 Feb 2004
Posts: 28
Location: sundsvall, sweden

PostPosted: Fri Feb 27, 2004 9:29 am    Post subject: Reply with quote

Many posts here is like "look how big my d**k is" so i really need to show u guys mine ;) I have cheap hitachi/ibm disks with only 2 mb cachememory, im surprised to see that they beats much newer and more expensive disks.

This is my system disk (hitachi/ibm 80gb)
Code:

root@gentoo wolfpakk # hdparm -tT /dev/hda

/dev/hda:
 Timing buffer-cache reads:   1008 MB in  2.01 seconds = 502.32 MB/sec
 Timing buffered disk reads:  168 MB in  3.01 seconds =  55.86 MB/sec



This is my second movies/mp3 disk (hitachi/ibm 120 gb)

Code:

root@gentoo wolfpakk # hdparm -tT /dev/hdc

/dev/hdc:
 Timing buffer-cache reads:   1012 MB in  2.00 seconds = 506.08 MB/sec
 Timing buffered disk reads:  162 MB in  3.04 seconds =  53.32 MB/sec
Back to top
View user's profile Send private message
abachn
n00b
n00b


Joined: 27 Feb 2004
Posts: 4
Location: Aarhus, Denmark

PostPosted: Fri Feb 27, 2004 10:52 am    Post subject: Reply with quote

I ran this on my P4 2.8 HT mashine and got these results. The disk is a 60 Gb maxtor disk.

/dev/hda:
Timing buffer-cache reads: 2696 MB in 2.00 seconds = 1346.86 MB/sec
Timing buffered disk reads: 154 MB in 3.04 seconds = 50.73 MB/sec
Back to top
View user's profile Send private message
wsmc884
n00b
n00b


Joined: 19 Nov 2003
Posts: 29

PostPosted: Sat Feb 28, 2004 6:07 am    Post subject: Reply with quote

I've been trying for some time now to get my new SATA drive to perform better without success. Here are the stats:

/dev/hde:
Timing buffer-cache reads: 1984 MB in 2.00 seconds = 990.67 MB/sec
Timing buffered disk reads: 82 MB in 3.06 seconds = 26.84 MB/sec

/dev/hde:

Model=ST3120023AS, FwRev=3.01, SerialNo=3KA1F0LN
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=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=234441648
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 udma3 udma4 udma5 *udma6
AdvancedPM=no WriteCache=enabled
Drive conforms to: ATA/ATAPI-6 T13 1410D revision 2:

* signifies the current active mode

/dev/hde:
multcount = 16 (on)
IO_support = 1 (32-bit)
unmaskirq = 1 (on)
using_dma = 1 (on)
keepsettings = 1 (on)
readonly = 0 (off)
readahead = 256 (on)
geometry = 65535/16/63, sectors = 234441648, start = 0

I've read this entire thread and several others over the past 8 months or so but I still can't get the disk reads above 30.

I'm using kernel 2.6.3

Any suggestions?
Back to top
View user's profile Send private message
sharky
n00b
n00b


Joined: 27 Jun 2002
Posts: 8
Location: Middelbeers, Netherlands

PostPosted: Sat Feb 28, 2004 3:16 pm    Post subject: Reply with quote

settings:

/dev/hda:
multcount = 16 (on)
IO_support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 1 (on)
keepsettings = 0 (off)
readonly = 0 (off)
readahead = 256 (on)
geometry = 16383/255/63, sectors = 117210240, start = 0

results:

/dev/hda:
Timing buffer-cache reads: 2588 MB in 2.00 seconds = 1294.20 MB/sec
Timing buffered disk reads: 112 MB in 3.00 seconds = 37.29 MB/sec

info:

/dev/hda:

Model=HTS726060M9AT00, FwRev=MH4OA68A, SerialNo=MRH400M4G2786A
Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs }
RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4
BuffType=DualPortCache, BuffSize=7877kB, MaxMultSect=16, MultSect=16
CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=117210240
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 udma3 udma4 *udma5
AdvancedPM=yes: mode=0x80 (128) WriteCache=enabled
Drive conforms to: ATA/ATAPI-6 T13 1410D revision 3a:

Machine info:

Laptop powered by a P4 3200 MHz, hyperthreading enabled, 1024 MB RAM DDR333, 60 GB 7200 rpm Hitachi HDD.
Back to top
View user's profile Send private message
RJARRRPCGP
n00b
n00b


Joined: 15 Feb 2004
Posts: 23
Location: USA (Springfield, Vermont)

PostPosted: Sun Feb 29, 2004 5:55 pm    Post subject: Reply with quote

My HDD performance results with Mandrake Linux 8.1 with Athlon T-bird 900 mhz, 256 MB PC133 SDRAM and Via KT133 chipset:

Timing buffer-cache reads: 128MB in 0.75 seconds =170.67 MB/sec

Timing buffered disk reads: 64 MB in 2.89 seconds = 22.15 MB/sec


The HDD is an IBM Deskstar 13 GB UDMA 66 7200 HDD (The reliable Deskstars before the new ones).

Is the benchmark bad?
Back to top
View user's profile Send private message
ck42
l33t
l33t


Joined: 31 Jul 2003
Posts: 789

PostPosted: Wed Mar 03, 2004 3:12 am    Post subject: Reply with quote

I'm a little stumped by the performance numbers I'm seeing. Judging from many of the (what I *think*) slower drives I see people posting, I thought my performance would be higher than it is.

I think I pretty much have the params maxed out.

-X69 -u1 -c1 -d1 -A1 -D1 -m16 -a8

/dev/hda:
multcount = 16 (on)
IO_support = 1 (32-bit)
unmaskirq = 1 (on)
using_dma = 1 (on)
keepsettings = 1 (on)
readonly = 0 (off)
readahead = 8 (on)
geometry = 65535/16/63, sectors = 160086528, start = 0
Timing buffered disk reads: 92 MB in 3.05 seconds = 30.13 MB/sec

This is with a Maxtor Diamondmax Plus 9 (7200RPM - 8mb cache).
hdparm shows that udma6 is supported but I don't think my m/b supports ata133 (if that is what udma6 is).

Am I missing anything here?


Last edited by ck42 on Wed Mar 03, 2004 9:23 pm; edited 1 time in total
Back to top
View user's profile Send private message
woody77
n00b
n00b


Joined: 06 Feb 2004
Posts: 69
Location: SC Mtns, CA, USA

PostPosted: Wed Mar 03, 2004 8:00 pm    Post subject: Reply with quote

So, what, exactly, ARE the bottlenecks that we can run into, and what are the workarounds?

1) Disk's own read-rate. Obviously, we can't read from disk any faster than it can read from the platters, so the max sequential read throughput of the disk is going to be a bottleneck for disk reads. This can be worked around with faster drives, but that's about it.

2) HD to controller. Next up seems to be maxing out the interface itself, (ata66/100/133, SATA150, etc.). That's a hard-limit, but can be worked around with RAID systems, or upgrading the MB/adapter cards/drives/cables.

3) The MB's own transfer rates between the controller and the processor. This is where it starts to get tricky. For a controller in the Southbridge, this is dictated by the chipset, and should be much, much, much higher than any of the others. However, if you've got a PCI-based controller, then the bottleneck is the PCI subsystem. And that's 33Mhz at 16 bits? or what, exactly?

From the reading on this thread, the best the hdparm can do is get the I/O subsystem running such that the bottleneck isn't the software, and how the data is being read, but instead removes all but the hardware bottlenecks that I've described above.

I can't seem to get more than 24MB/sec on my laptop. This being a P4m, with an i845 chipset, the bottleneck is most likely the drive (ATA100/udma5), and not anything else. It's a 5400rpm drive, but the heads are slow, so seeks are horridly slow (24ms), but the sequential read-rate also seems limited.

Those of you seeing much higher transfer rates, I'm guessing that you've got the SATA or udma6, and have drives that can take better advantage of those supporting systems. However, it seems like bothering with the high-speed interfaces when the drives themselves can't push the data is a little useless.

SCSI is different, where you've got multiple drives on the chain, and needing 3 drives to max out the chain actually makes RAID give some BIG performance boosts (ie, RAID0 across three drives could potentially give you full SCSI saturation, and 160MB/320MB transfer rates).

I guess it depends on the cache, doesn't it? I've got a 16MB cache, so if you're doing burst reads, you get full saturation of the pipe, but if doing large sequential reads, you can't saturate the pipe because the drive can't read the data off fast enough.

And then comes the question, what are you trying to optimize for... And that's a whole 'nother ball of wax.
Back to top
View user's profile Send private message
der_maddin
Apprentice
Apprentice


Joined: 22 Feb 2004
Posts: 209
Location: somewhere between my systems ..

PostPosted: Thu Mar 04, 2004 5:43 pm    Post subject: Reply with quote

ck42 wrote:


I think I pretty much have the params maxed out.

-X69 -u1 -c1 -d1 -A1 -D1 -m16 -a8



i tried this settings on my harddrive,
( wich is equal to yours, its a samsung 160gb, 8M cache, 7200rpm )
they really slowed it down :
( see post above for original settings / speeds )

Code:

 Timing buffer-cache reads:   1024 MB in  2.00 seconds = 510.80 MB/sec
 Timing buffered disk reads:   84 MB in  3.01 seconds =  27.91 MB/sec


this seems equal to the values you got.
but when played around a bit with the params even
more than in my earlier post i get this :

Code:

 Timing buffer-cache reads:   1028 MB in  2.00 seconds = 513.05 MB/sec
 Timing buffered disk reads:  170 MB in  3.03 seconds =  56.11 MB/sec


this was achieved using
hdparm -X69 -u1 -c3 -d1 -A1 -m16 -a256 -k1
( don't forget the keep settings - flag '-k1' )

i would recommend you to try these settings for yourself
and then report back to this thread.

--------------------------------------------------

woody77 wrote:

I've got a 16MB cache


are you serious ?? on a notebook system ?
i have never heard of a harddisks cache that
big not even on desktop hdds and you're talking
about a notebook harddrive ??
is that possible ?

------------------------------------------------

now something quite different ;)

i do still wonder why i can't select udma6 for my harddrive.
the motherboards chipset ( a sis746 ) does support it,
so does the harddrive itself, i even use a 133 cable.
so why cant i ?
_________________
neoCortex maddin # uname -a
Linux neoCortex 2.6.14-gentoo-r5 #2 SMP Fri Jan 13 04:40:02 CET 2006 x86_64 AMD Athlon(tm) 64 Processor 3400+ AuthenticAMD GNU/Linux
Back to top
View user's profile Send private message
ck42
l33t
l33t


Joined: 31 Jul 2003
Posts: 789

PostPosted: Thu Mar 04, 2004 6:48 pm    Post subject: Reply with quote

ker_maddin:

I'm using the -a8 setting because that's what hdparm reports that the drive supports. Have you verified that your's supports the 256 setting?

Also, I was originally using the c3 option but from what I understand, c1 seems more performance oriented. I've even tried testing using both and there's a SMALL SMALL improvement using c3.

..oh, and yes, I use the k1 setting. Just forgot to post that when I typed in the msg.


EDIT:

Went ahead and tried the -a 256 setting.
Things DEF. improved.
Timing buffered disk reads: 162 MB in 3.02 seconds = 53.62 MB/sec

Ran this a couple of times using both c1 and c3. Looks like the c3 option is ever so slightly faster (~1Mb/sec)...beating out the c1 setting.

thnx for the tip!


Last edited by ck42 on Thu Mar 04, 2004 8:22 pm; edited 1 time in total
Back to top
View user's profile Send private message
woody77
n00b
n00b


Joined: 06 Feb 2004
Posts: 69
Location: SC Mtns, CA, USA

PostPosted: Thu Mar 04, 2004 8:12 pm    Post subject: Reply with quote

der_maddin wrote:

woody77 wrote:

I've got a 16MB cache


are you serious ?? on a notebook system ?
i have never heard of a harddisks cache that
big not even on desktop hdds and you're talking
about a notebook harddrive ??
is that possible ?


Well, of course it's POSSIBLE, anything like that is. However, it is true. Blew my mind when I found out. I figured that the 8MB on the IBM Travelstars that it's a knockoff of was huge (40GNX?). But it's a great drive, aside from coming from the factory with the write-cache enabled, and XP defaulting to enabled (I find it scary, but no data-loss yet, and everything I do is backed up on servers in CVS).

BTW, it was the supplied drive in my IBM Thinkpad T30. I was definitely amused to find out it was a Toshiba knockoff of the Hitachi/IBM drive...

It's a Toshiba MK-4019GAX. specs or here and go to tech specs:

Code:

Basic Specifications & Features
   Part Number
   HDD2171*ZE
   Formatted Capacity
   40,0078 Mbyte
   Form Factor
   2.5 Inch
   Interface Type
#    ATAPI/IDE
   Interface Standards supported
#    ATA-2
# ATA-3
# ATA-4
# ATA-5
   Head/Actuator Technology
#    "Ramp-Load" Technology
# GMR Head
   Physical Parameters
      Number of Cylinders
   33,440
      Number of Heads
   4
   Number of Platters
   2
   Bytes/Sector
   512
   Shock Detection
   Shock sensor circuit
   SMART
   The SMART command set is supported.
Performance Specifications
   Average Seek Time
   12 ms
   Maximum Seek Time
   22 ms
   Track-To-Track Seek Time
   2 ms
   Internal Transfer Rate
   From 200.8 to 333.2 Mbit/s
   Host Transfer Rate
      PIO Mode
#    PIO Mode 4 (16.6 Mbyte/s)
      UDMA Mode
#    UDMA Mode 5/100 (100 MByte/s)
   Rotational Speed
   5,400 rpm
   Buffer Size
   16,384 Kbyte
   Cache
#    Read-Ahead Cache
# Write Cache


Last edited by woody77 on Thu Mar 04, 2004 8:23 pm; edited 1 time in total
Back to top
View user's profile Send private message
der_maddin
Apprentice
Apprentice


Joined: 22 Feb 2004
Posts: 209
Location: somewhere between my systems ..

PostPosted: Thu Mar 04, 2004 8:22 pm    Post subject: Reply with quote

at wich point does hdparm say your device only supports 8 sectors of filesystem read-ahead ??

i've tried values from 32 - 8192 ( means 16 - 4096 kB )
without hdparm arguing about, but also without any
performance gain.
so far, i do understand that the max value for the -a
parameter could be the complete cache capacity,
that means for our devices a value of
16384, ( 2*8192 ) as they have 8M of cache.

see this excerpt from man hdparm :

Code:

In  the  current  kernel  version
(2.0.10)  this  has  a default setting of 8 sectors (4KB).


did you see it talks about 'current' kernel version 2.0.10 ;)
and its the most recent version of hdparm ( 5.5 ) i'm using !!

and the difference between -c1 and -c3 is also described in the
hparm man file, it says that they are almost the same, only
that -c3 uses some kind of sync sequence wich is required
for most chipsets, but both of them use 32 bit I/O.

btw, the hdparm -t speeds are only reference values,
the actual speeds may vary depending on the actual
size of sequential streams you read.
_________________
neoCortex maddin # uname -a
Linux neoCortex 2.6.14-gentoo-r5 #2 SMP Fri Jan 13 04:40:02 CET 2006 x86_64 AMD Athlon(tm) 64 Processor 3400+ AuthenticAMD GNU/Linux
Back to top
View user's profile Send private message
ck42
l33t
l33t


Joined: 31 Jul 2003
Posts: 789

PostPosted: Thu Mar 04, 2004 8:31 pm    Post subject: Reply with quote

der_maddin:

Can't find where I found that info (-i or I) but I coulda sworn I saw it somewhere.

Other question is regarding using UDMA6. The drive itself is an ATA133 drive but the m/b only supports ATA100 (I'll have to double check to be sure).

The info section shows that UDMA6 is available....but I'm not sure I want to set this if it's something that the m/b doesn't support.

EDIT: woops....looks like my m/b DOES support ATA133!! time to test it :twisted:

OK.....no difference/improvment. :roll:


Last edited by ck42 on Thu Mar 04, 2004 8:35 pm; edited 1 time in total
Back to top
View user's profile Send private message
woody77
n00b
n00b


Joined: 06 Feb 2004
Posts: 69
Location: SC Mtns, CA, USA

PostPosted: Thu Mar 04, 2004 8:33 pm    Post subject: Reply with quote

der_maddin wrote:
where does hdparm say your device only supports 8 sectors of filesystem
read-ahead ??


I was about to say that it's rejected past 8, but that was multi-sector reads (-m8).

Quote:

i've tried values from 32 - 8192 ( means 16 - 4096 kB )
without hdparm arguing about, but also without any
performance gain.
so far, i do understand that the max value for the -a
parameter could be the complete cache capacity,
that means for our devices a value of
16384, ( 2*8192 ) as they have 8M of cache.


Hmm. worth playing with, but IIRC, I didn't see any performance gains past 15 or so. But in those cases, I *think* I'm running into the limit at which the heads can read the data off the drive. 200-333Mb/sec works out to 25-41MB/sec. However, my Linux partition is at the very end of the disk, where I know the transfer rates aren't as good. So that might be why only 25MB or so on my drive.
Back to top
View user's profile Send private message
der_maddin
Apprentice
Apprentice


Joined: 22 Feb 2004
Posts: 209
Location: somewhere between my systems ..

PostPosted: Thu Mar 04, 2004 8:42 pm    Post subject: Reply with quote

wow, woody, i wont believe this if it wasn't true...

now you made me curious, as i own a toshiba notebook
( Satellite Pro 6100) myself.
it uses a hitachi harddrive DK23EA-60
that only has 2M of cache :(

you are a lucky one ..

EDIT : i just looked your drive up at a local seller,
( because i am now somewhat jealous )
and found out your hdd model ( MK-4026GAX )
is listed for 139.00 euros and another bigger
one ( 80 GB / MK8026GAX ) is available for
279.00.

Perhaps i buy one of these so i can match up with
your specs, just because i'm really jealous right
now...

btw, what notebook model are you running the disk in ??
_________________
neoCortex maddin # uname -a
Linux neoCortex 2.6.14-gentoo-r5 #2 SMP Fri Jan 13 04:40:02 CET 2006 x86_64 AMD Athlon(tm) 64 Processor 3400+ AuthenticAMD GNU/Linux
Back to top
View user's profile Send private message
der_maddin
Apprentice
Apprentice


Joined: 22 Feb 2004
Posts: 209
Location: somewhere between my systems ..

PostPosted: Thu Mar 04, 2004 8:45 pm    Post subject: Reply with quote

ck42 wrote:

Other question is regarding using UDMA6. The drive itself is an ATA133 drive but the m/b only supports ATA100 (I'll have to double check to be sure).

The info section shows that UDMA6 is available....but I'm not sure I want to set this if it's something that the m/b doesn't support.

EDIT: woops....looks like my m/b DOES support ATA133!! time to test it :twisted:

OK.....no difference/improvment. :roll:


lucky one.
i don't have so much luck with that.
both my hdd and my mobo do support ATA133 but
somehow udma6 is not available for me ... :(
_________________
neoCortex maddin # uname -a
Linux neoCortex 2.6.14-gentoo-r5 #2 SMP Fri Jan 13 04:40:02 CET 2006 x86_64 AMD Athlon(tm) 64 Processor 3400+ AuthenticAMD GNU/Linux
Back to top
View user's profile Send private message
woody77
n00b
n00b


Joined: 06 Feb 2004
Posts: 69
Location: SC Mtns, CA, USA

PostPosted: Thu Mar 04, 2004 9:38 pm    Post subject: Reply with quote

You must have missed my edit where I added it's an IBM T30. NICE laptop.

EU$130? Not bad. Have to check around here for prices. Want a new drive in the other laptop (9GB, slow, and LOUD). The GAX is silent. audible only when seeking.

UDMA/133 is only going to get better burst rates, any sequential stuff is probably limited by the drive. Lots of smallish reads (ie, fully from the cache) will be faster on UDMA/133, but the drive itself is the bottleneck.

The internal media transfer rate, platter rpm, and seek times are probably MUCH more important than the bus, as a single drive can't suturate the bus.

i.e., my drive is FAST, when it's all out of the cache, but if you need to go to disk, it's SLOW... 25-40MB/sec from the platters, vs. 100MB/sec from disk cache to the host.
Back to top
View user's profile Send private message
der_maddin
Apprentice
Apprentice


Joined: 22 Feb 2004
Posts: 209
Location: somewhere between my systems ..

PostPosted: Thu Mar 04, 2004 10:02 pm    Post subject: Reply with quote

woooh.
i just found some review on the t30.
saw it would cost $3,499 ( in june 2002 )
thats pretty much.

but i'm sure its worth it;
as far as i know from friends also owning an ibm
nb, those seem to be perfectly designed.

i paid about 2,400 ? for my Satellite
in jan of last year. it has the pentium-M 2.4
and a geForce 4go, 60 GB harddisk.
but it lacks a touchpad.

i use it mostly for work, that
also explains why it is the fastest system
i own ( my desktop is running 'only' an athlonXP 1900+ )
also its the only system i have running
windows ( 2000 ).
_________________
neoCortex maddin # uname -a
Linux neoCortex 2.6.14-gentoo-r5 #2 SMP Fri Jan 13 04:40:02 CET 2006 x86_64 AMD Athlon(tm) 64 Processor 3400+ AuthenticAMD GNU/Linux
Back to top
View user's profile Send private message
woody77
n00b
n00b


Joined: 06 Feb 2004
Posts: 69
Location: SC Mtns, CA, USA

PostPosted: Thu Mar 04, 2004 11:04 pm    Post subject: Reply with quote

It's a work laptop. Home laptop is a Dell. Not going to buy another Dell laptop. It's WAY too flimsy. IBM laptops are expensive, but good for people who travel on business. They're heavy, but they're solid. I'll take the rigidity and strength of the IBM frames over the flimsy dell ones.

But yeah, they're expensive. Fast, but expensive.
Back to top
View user's profile Send private message
der_maddin
Apprentice
Apprentice


Joined: 22 Feb 2004
Posts: 209
Location: somewhere between my systems ..

PostPosted: Fri Mar 05, 2004 1:24 am    Post subject: Reply with quote

i also mispicked a notebook once.
it was a gericom model '1st supersonic'
( first model on market with 1GHz )
that thing really sucked ...

i had it repaired three times
everytime because of the same flaw.

it crashed unpredictable and the
battery was defective after 3~5
weeks, it was replaced all three times,
but same shit everytime..

worst of all this thing was
expensive :?
_________________
neoCortex maddin # uname -a
Linux neoCortex 2.6.14-gentoo-r5 #2 SMP Fri Jan 13 04:40:02 CET 2006 x86_64 AMD Athlon(tm) 64 Processor 3400+ AuthenticAMD GNU/Linux
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware All times are GMT
Goto page Previous  1, 2, 3, 4, 5, 6, 7, 8, 9  Next
Page 6 of 9

 
Jump to:  
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