hard drive problem

"Marc" <marcoraicevic.TakeThisOut@videotron.ca> wrote in message
news:EAH8j.13432$Mg3.28076@weber.videotron.net...
>I just installed a second western digital 500Gb sata2 hard drive (used for
> storage and playing music) on my old non-sata supporting pc. I have an amd
> athlon xp 2000+1.67 Ghz processor, a foxcom motherboard which does not
> support sata, I have a sonata II case which supports sata, 768MB of ram,
> my
> main western digital 160Gb ide hard drive which is formated in fat 32,
> partitioned in three drives, and has windows xp os installed on the c
> drive.
> I installed a pci serial ata host controller card on a pci slot on my
> motherboard and connected the controller card to the sata hard drive. I
> formated the 500Gb sata drive in fat 32 with one primary partition using
> western digital data life guard tools. I ended up transferring around
> 160Gb
> of music files to this drive from my western digital 160Gb ide hard drive
> and have been using the drive to store and play music for around a month.
> But now when I try using the sata drive or playing music it slows my whole
> system down to a crawl, the system is unusuable. However, when I unplug
> the
> sata drive the rest of the system works fine. So I disconnected the sata
> drive and now my system with the ide hard drive - all of the drives have
> been defragmented - is working fine. I was suspecting that the reason the
> system was slow was because the 500Gb sata drive is fragmented. However, I
> don't know how to go about defragging the 500Gb sata hard drive since my
> defragger (diskeeper) is on the 160Gb ide hard drive - the drive which has
> my winxp operating system and diskeeper.
> Is there anyone that can give me a solution to this problem?
> Thanks,
> Marc
>
>
>
>
>
First thought is to run the WD's diagnostics program on your new hdd. Sounds
strange that only after a month the drive should be grinding to a crawl.
Second is that I would guess one could run the Windows disk defragmentor
on the new drive. You would find it under Accesories and System Tools.
--
Jan Alter
bearpuf.TakeThisOut@verizon.net
or
jalter.TakeThisOut@phila.k12.pa.us
 >> Stay informed about: hard drive problem 

I got diskeeper to recognize the sata drive. I'm defragging the sata drive.
I'm thinking of transferring some music files to my other drives and using
the ide drive to listen to music and I'll get some more ram. Perhaps the
combination of ripping several movies and playing mp3's took up too many
resources and caused the whole system to crash. Now that I disconnected and
reconnected the sata drive to the pci sata controller and rebooted it seems
to be working okay.

"Jan Alter" <bearpuf.DeleteThis@verizon.net> wrote in message
news:mTP8j.141$TZ4.112@trnddc02...
> "Marc" <marcoraicevic.DeleteThis@videotron.ca> wrote in message
> news:EAH8j.13432$Mg3.28076@weber.videotron.net...
>>I just installed a second western digital 500Gb sata2 hard drive (used for
>> storage and playing music) on my old non-sata supporting pc. I have an
>> amd
>> athlon xp 2000+1.67 Ghz processor, a foxcom motherboard which does not
>> support sata, I have a sonata II case which supports sata, 768MB of ram,
>> my
>> main western digital 160Gb ide hard drive which is formated in fat 32,
>> partitioned in three drives, and has windows xp os installed on the c
>> drive.
>> I installed a pci serial ata host controller card on a pci slot on my
>> motherboard and connected the controller card to the sata hard drive. I
>> formated the 500Gb sata drive in fat 32 with one primary partition using
>> western digital data life guard tools. I ended up transferring around
>> 160Gb
>> of music files to this drive from my western digital 160Gb ide hard drive
>> and have been using the drive to store and play music for around a month.
>> But now when I try using the sata drive or playing music it slows my
>> whole
>> system down to a crawl, the system is unusuable. However, when I unplug
>> the
>> sata drive the rest of the system works fine. So I disconnected the sata
>> drive and now my system with the ide hard drive - all of the drives have
>> been defragmented - is working fine. I was suspecting that the reason the
>> system was slow was because the 500Gb sata drive is fragmented. However,
>> I
>> don't know how to go about defragging the 500Gb sata hard drive since my
>> defragger (diskeeper) is on the 160Gb ide hard drive - the drive which
>> has
>> my winxp operating system and diskeeper.
>> Is there anyone that can give me a solution to this problem?
>> Thanks,
>> Marc
>>
>>
>>
>>
>>
> First thought is to run the WD's diagnostics program on your new hdd.
> Sounds strange that only after a month the drive should be grinding to a
> crawl.
> Second is that I would guess one could run the Windows disk defragmentor
> on the new drive. You would find it under Accesories and System Tools.
> --
> Jan Alter
> bearpuf.DeleteThis@verizon.net
> or
> jalter.DeleteThis@phila.k12.pa.us
>
 >> Stay informed about: hard drive problem 

On 2007-12-15, kony <spam.RemoveThis@spam.com> wrote:
> On Fri, 14 Dec 2007 22:10:28 -0500, "Marc"
><marcoraicevic.RemoveThis@videotron.ca> wrote:
>
>>I installed a pci serial ata host controller card on a pci slot on my
>>motherboard and connected the controller card to the sata hard drive. I
>>formated the 500Gb sata drive in fat 32 with one primary partition using
>>western digital data life guard tools. I ended up transferring around 160Gb
>>of music files to this drive from my western digital 160Gb ide hard drive
>>and have been using the drive to store and play music for around a month.
>>But now when I try using the sata drive or playing music it slows my whole
>>system down to a crawl, the system is unusuable.
>
> What has changed since the point where it worked ok, and now
> when it doesn't? It seems most likely a software problem,
> BUT check the cables including unplugging and replugging
> them for the drive.

Is there any reason for formatting the drive as FAT32? FAT32 _is_
slow on large hard drives - there is a reason why Microsoft's tools
won't format large FAT32 partitions. I'd suggest retrying the
drive formatted as NTFS.

The primary difficulty with FAT32 is locating free space on the
drive: for small partitions it isn't so much of a problem as there
is less data to sort through, but larger partitions become
proportionately slower. The speed will decrease over time in the
manner you describe as the drive fills up and the free space becomes
fragmented.

--
Andrew Smallshaw
andrews.RemoveThis@sdf.lonestar.org
 >> Stay informed about: hard drive problem 

On Sat, 15 Dec 2007 21:41:05 +0100 (CET), Andrew Smallshaw
<andrews.DeleteThis@sdf.lonestar.org> wrote:

>On 2007-12-15, kony <spam.DeleteThis@spam.com> wrote:
>> On Fri, 14 Dec 2007 22:10:28 -0500, "Marc"
>><marcoraicevic.DeleteThis@videotron.ca> wrote:
>>
>>>I installed a pci serial ata host controller card on a pci slot on my
>>>motherboard and connected the controller card to the sata hard drive. I
>>>formated the 500Gb sata drive in fat 32 with one primary partition using
>>>western digital data life guard tools. I ended up transferring around 160Gb
>>>of music files to this drive from my western digital 160Gb ide hard drive
>>>and have been using the drive to store and play music for around a month.
>>>But now when I try using the sata drive or playing music it slows my whole
>>>system down to a crawl, the system is unusuable.
>>
>> What has changed since the point where it worked ok, and now
>> when it doesn't? It seems most likely a software problem,
>> BUT check the cables including unplugging and replugging
>> them for the drive.
>
>Is there any reason for formatting the drive as FAT32? FAT32 _is_
>slow on large hard drives - there is a reason why Microsoft's tools
>won't format large FAT32 partitions. I'd suggest retrying the
>drive formatted as NTFS.

Nope, FAT32 is not much slower. It will not account for
this, you would seldom if ever notice a difference in use
unless there was a lot of small files moving back and forth
over a long period to fragment it, and otherwise only
certain types of benchmarks (and even then FAT32 sometimes
being faster than NTFS). Besides this, the data rate for
playing music is sufficiently low that it's just not
possible to be caused by FAT32 fragmentation or other FAT32
issues... though I suppose if something were corrupted then
remaking the FAT32 (or NTFS) partition would be a good thing
to try.

Further, general drive tests like checkdisk and the HDD
manufacturer's diagnostics, and trying a different data
cable would also be good things to consider.

Microsoft's tools won't format larger FAT32 partitions for
only one reason - they want to make the industry as
proprietary and MS-centric as possible.

It's not that I'm against NTFS, it certainly has some
advantages like security and > 4GB file support, but
performance is not a realistic factor. Then there's the
downside, limited to non-existant support of NTFS on a
non-*NT* based system.


>
>The primary difficulty with FAT32 is locating free space on the
>drive: for small partitions it isn't so much of a problem as there
>is less data to sort through, but larger partitions become
>proportionately slower. The speed will decrease over time in the
>manner you describe as the drive fills up and the free space becomes
>fragmented.

NTFS fragments. FAT32 can be defragged. Multi-MB music
files aren't going to substantially fragment, they are
generally written whole and don't fragment at all.

The problem just isn't FAT32. It almost never is, those who
pick one or the other for performance reasons on a PC are
mostly delusional.
 >> Stay informed about: hard drive problem 

"Andrew Smallshaw" <andrews RemoveThis @sdf.lonestar.org> wrote in message
news:slrnfm8euq.s60.andrews@sdf.lonestar.org...
> On 2007-12-15, kony <spam RemoveThis @spam.com> wrote:
>> On Fri, 14 Dec 2007 22:10:28 -0500, "Marc"
>><marcoraicevic RemoveThis @videotron.ca> wrote:
>>
>>>I installed a pci serial ata host controller card on a pci slot on my
>>>motherboard and connected the controller card to the sata hard drive. I
>>>formated the 500Gb sata drive in fat 32 with one primary partition using
>>>western digital data life guard tools. I ended up transferring around
>>>160Gb
>>>of music files to this drive from my western digital 160Gb ide hard drive
>>>and have been using the drive to store and play music for around a month.
>>>But now when I try using the sata drive or playing music it slows my
>>>whole
>>>system down to a crawl, the system is unusuable.
>>
>> What has changed since the point where it worked ok, and now
>> when it doesn't? It seems most likely a software problem,
>> BUT check the cables including unplugging and replugging
>> them for the drive.
>
> Is there any reason for formatting the drive as FAT32? FAT32 _is_
> slow on large hard drives - there is a reason why Microsoft's tools
> won't format large FAT32 partitions. I'd suggest retrying the
> drive formatted as NTFS.
>
> The primary difficulty with FAT32 is locating free space on the
> drive: for small partitions it isn't so much of a problem as there
> is less data to sort through, but larger partitions become
> proportionately slower. The speed will decrease over time in the
> manner you describe as the drive fills up and the free space becomes
> fragmented.
>
> --
> Andrew Smallshaw
> andrews RemoveThis @sdf.lonestar.org

I guess I could get a third storage hard drive and format it in ntfs. There
would not
be any problems using my main hard drive and the second hard drive both
formated
in fat 32 with a third drive formated in ntfs would there? The problem was
not
because defragmentation, since the second 500Gb sata hard drive was hardly
defragmented according to dsikeeper. The problem was probably due
to some software glitch due to using too many resouces ripping, playing
music,
browsing, and playing online games.
 >> Stay informed about: hard drive problem 

"Jan Alter" <bearpuf.DeleteThis@verizon.net> wrote in message
news:mTP8j.141$TZ4.112@trnddc02...
> "Marc" <marcoraicevic.DeleteThis@videotron.ca> wrote in message
> news:EAH8j.13432$Mg3.28076@weber.videotron.net...
>>I just installed a second western digital 500Gb sata2 hard drive (used for
>> storage and playing music) on my old non-sata supporting pc. I have an
>> amd
>> athlon xp 2000+1.67 Ghz processor, a foxcom motherboard which does not
>> support sata, I have a sonata II case which supports sata, 768MB of ram,
>> my
>> main western digital 160Gb ide hard drive which is formated in fat 32,
>> partitioned in three drives, and has windows xp os installed on the c
>> drive.
>> I installed a pci serial ata host controller card on a pci slot on my
>> motherboard and connected the controller card to the sata hard drive. I
>> formated the 500Gb sata drive in fat 32 with one primary partition using
>> western digital data life guard tools. I ended up transferring around
>> 160Gb
>> of music files to this drive from my western digital 160Gb ide hard drive
>> and have been using the drive to store and play music for around a month.
>> But now when I try using the sata drive or playing music it slows my
>> whole
>> system down to a crawl, the system is unusuable. However, when I unplug
>> the
>> sata drive the rest of the system works fine. So I disconnected the sata
>> drive and now my system with the ide hard drive - all of the drives have
>> been defragmented - is working fine. I was suspecting that the reason the
>> system was slow was because the 500Gb sata drive is fragmented. However,
>> I
>> don't know how to go about defragging the 500Gb sata hard drive since my
>> defragger (diskeeper) is on the 160Gb ide hard drive - the drive which
>> has
>> my winxp operating system and diskeeper.
>> Is there anyone that can give me a solution to this problem?
>> Thanks,
>> Marc
>>
>>
>>
>>
>>
> First thought is to run the WD's diagnostics program on your new hdd.
> Sounds strange that only after a month the drive should be grinding to a
> crawl.
> Second is that I would guess one could run the Windows disk defragmentor
> on the new drive. You would find it under Accesories and System Tools.
> --
> Jan Alter
> bearpuf.DeleteThis@verizon.net
> or
> jalter.DeleteThis@phila.k12.pa.us
>

I got diskeeper to work on the sata drive. After I unplugged and reconnected
the sata drive to the controller everything seems to be working fine now.
 >> Stay informed about: hard drive problem 

"kony" <spam.DeleteThis@spam.com> wrote in message
news:cul8m31doov7k2nlksbk1q3v4i1e49msr2@4ax.com...
> On Sat, 15 Dec 2007 21:41:05 +0100 (CET), Andrew Smallshaw
> <andrews.DeleteThis@sdf.lonestar.org> wrote:
>
>>On 2007-12-15, kony <spam.DeleteThis@spam.com> wrote:
>>> On Fri, 14 Dec 2007 22:10:28 -0500, "Marc"
>>><marcoraicevic.DeleteThis@videotron.ca> wrote:
>>>
>>>>I installed a pci serial ata host controller card on a pci slot on my
>>>>motherboard and connected the controller card to the sata hard drive. I
>>>>formated the 500Gb sata drive in fat 32 with one primary partition using
>>>>western digital data life guard tools. I ended up transferring around
>>>>160Gb
>>>>of music files to this drive from my western digital 160Gb ide hard
>>>>drive
>>>>and have been using the drive to store and play music for around a
>>>>month.
>>>>But now when I try using the sata drive or playing music it slows my
>>>>whole
>>>>system down to a crawl, the system is unusuable.
>>>
>>> What has changed since the point where it worked ok, and now
>>> when it doesn't? It seems most likely a software problem,
>>> BUT check the cables including unplugging and replugging
>>> them for the drive.
>>
>>Is there any reason for formatting the drive as FAT32? FAT32 _is_
>>slow on large hard drives - there is a reason why Microsoft's tools
>>won't format large FAT32 partitions. I'd suggest retrying the
>>drive formatted as NTFS.
>
> Nope, FAT32 is not much slower. It will not account for
> this, you would seldom if ever notice a difference in use
> unless there was a lot of small files moving back and forth
> over a long period to fragment it, and otherwise only
> certain types of benchmarks (and even then FAT32 sometimes
> being faster than NTFS). Besides this, the data rate for
> playing music is sufficiently low that it's just not
> possible to be caused by FAT32 fragmentation or other FAT32
> issues... though I suppose if something were corrupted then
> remaking the FAT32 (or NTFS) partition would be a good thing
> to try.
>
> Further, general drive tests like checkdisk and the HDD
> manufacturer's diagnostics, and trying a different data
> cable would also be good things to consider.
>
> Microsoft's tools won't format larger FAT32 partitions for
> only one reason - they want to make the industry as
> proprietary and MS-centric as possible.
>
> It's not that I'm against NTFS, it certainly has some
> advantages like security and > 4GB file support, but
> performance is not a realistic factor. Then there's the
> downside, limited to non-existant support of NTFS on a
> non-*NT* based system.
>
>
>>
>>The primary difficulty with FAT32 is locating free space on the
>>drive: for small partitions it isn't so much of a problem as there
>>is less data to sort through, but larger partitions become
>>proportionately slower. The speed will decrease over time in the
>>manner you describe as the drive fills up and the free space becomes
>>fragmented.
>
> NTFS fragments. FAT32 can be defragged. Multi-MB music
> files aren't going to substantially fragment, they are
> generally written whole and don't fragment at all.
>
> The problem just isn't FAT32. It almost never is, those who
> pick one or the other for performance reasons on a PC are
> mostly delusional.
>

The problem is gone since I reconnected the sata drive to the controller.
 >> Stay informed about: hard drive problem 

On 2007-12-15, kony <spam RemoveThis @spam.com> wrote:
> On Sat, 15 Dec 2007 21:41:05 +0100 (CET), Andrew Smallshaw
><andrews RemoveThis @sdf.lonestar.org> wrote:
>>
>>Is there any reason for formatting the drive as FAT32? FAT32 _is_
>>slow on large hard drives - there is a reason why Microsoft's tools
>>won't format large FAT32 partitions. I'd suggest retrying the
>>drive formatted as NTFS.
>
> Nope, FAT32 is not much slower. It will not account for
> this, you would seldom if ever notice a difference in use
> unless there was a lot of small files moving back and forth
> over a long period to fragment it, and otherwise only
> certain types of benchmarks (and even then FAT32 sometimes
> being faster than NTFS). Besides this, the data rate for

Have you actually looked at the FAT32 data structures? Where is
the free list or free space bitmap? There isn't one. Finding a
free sector to write to involves a linear scan of the FAT looking
for an unallocated cluster. Why have Microsoft introduced exFAT,
which addresses this issue amongst others, for large volumes where
the overhead of NTFS is not justified?

FAT's fundamental limitations are fine on small to medium sized
volumes, but they become significant on larger drives as the FAT
itself becomes correspondingly larger. NTFS has its own overheads,
of course, but these are due it it being a robust, modern filesystem
design. At worst the overhead varies logarithmically with filesystem
size, not linearly as is the case with any FAT with the same cluster
size.

> Microsoft's tools won't format larger FAT32 partitions for
> only one reason - they want to make the industry as
> proprietary and MS-centric as possible.

Microsoft have claimed several patents in the US regarding FAT32
(they probably wouldn't be allowed to stand elsewhere). FAT32 is
already propreitary enough - I'm sure you could find reports of
the licensing terms on the net if you were so inclined. In any
case Microsoft have a technically sound justification for this
decision at http://tinyurl.com/353gcq which also touches on some
of the other points I have made.

> It's not that I'm against NTFS, it certainly has some
> advantages like security and > 4GB file support, but
> performance is not a realistic factor. Then there's the
> downside, limited to non-existant support of NTFS on a
> non-*NT* based system.

NTFS support on non-Windows platforms has improved greatly over
the last three years or so. It isn't really a problem anymore.
Hell, even NetBSD supports it now, and that tends to stay slightly
behind the curve in terms of its capabilities.

> NTFS fragments. FAT32 can be defragged. Multi-MB music
> files aren't going to substantially fragment, they are
> generally written whole and don't fragment at all.

NTFS doesn't fragment to anything like the extent FAT filesystems
do. Bitmapping of free space and keeping metadata seperated from
real data work wonders. Fragmentation with NTFS is only really an
issue if your disk is almost full or you particularly care about
the last 1% performance - it doesn't cripple a system as can happen
with FAT.

> The problem just isn't FAT32. It almost never is, those who
> pick one or the other for performance reasons on a PC are
> mostly delusional.

A lot of the time I would agree with you that the differences are
slight. However in this case the OP has used a third party tool
to create a file system over 15x larger than the cap Microsoft has
placed there for sound engineering reasons. Doesn't that ring any
alarm bells that maybe things are being pushed beyond a sensible
limit?

--
Andrew Smallshaw
andrews RemoveThis @sdf.lonestar.org
 >> Stay informed about: hard drive problem 

On 2007-12-15, Marc <marcoraicevic RemoveThis @videotron.ca> wrote:
>
> I guess I could get a third storage hard drive and format it in
> ntfs. There would not be any problems using my main hard drive and
> the second hard drive both formated in fat 32 with a third drive
> formated in ntfs would there?

That won't be a problem, you can mix match filesystem types any
way you want.

> The problem was not because defragmentation, since the second 500Gb
> sata hard drive was hardly defragmented according to dsikeeper.

If it is what I suspect then it is a different problem to what is
usually meant by disk fragmentation. If I am correct then it is
the FAT itself, not the data on disk, that is causing the problem.
A defragmenter may actually make the problem _worse_ since a full
defrag will pack all the data at the beginning of the partition.
That means that any writes have to pass over all of your exisiting
data before some free space can be found.

--
Andrew Smallshaw
andrews RemoveThis @sdf.lonestar.org
 >> Stay informed about: hard drive problem 

On Sun, 16 Dec 2007 21:56:16 +0100 (CET), Andrew Smallshaw
<andrews.RemoveThis@sdf.lonestar.org> wrote:

>On 2007-12-15, Marc <marcoraicevic.RemoveThis@videotron.ca> wrote:
>>
>> I guess I could get a third storage hard drive and format it in
>> ntfs. There would not be any problems using my main hard drive and
>> the second hard drive both formated in fat 32 with a third drive
>> formated in ntfs would there?
>
>That won't be a problem, you can mix match filesystem types any
>way you want.
>
>> The problem was not because defragmentation, since the second 500Gb
>> sata hard drive was hardly defragmented according to dsikeeper.
>
>If it is what I suspect then it is a different problem to what is
>usually meant by disk fragmentation. If I am correct then it is
>the FAT itself, not the data on disk, that is causing the problem.
>A defragmenter may actually make the problem _worse_ since a full
>defrag will pack all the data at the beginning of the partition.
>That means that any writes have to pass over all of your exisiting
>data before some free space can be found.


Hardly.
Seek time is mere milliseconds. These aren't files that
would fit in NTFS file table, the very thing you were
previously claiming a virtue of NTFS is now the thing you
try to use to condemn FAT32, that NTFS would also have
packed data and writies further in on the platter. Give it
up already, it wasn't a FAT32 problem.
 >> Stay informed about: hard drive problem 

On Sun, 16 Dec 2007 21:38:10 +0100 (CET), Andrew Smallshaw
<andrews DeleteThis @sdf.lonestar.org> wrote:

>On 2007-12-15, kony <spam DeleteThis @spam.com> wrote:
>> On Sat, 15 Dec 2007 21:41:05 +0100 (CET), Andrew Smallshaw
>><andrews DeleteThis @sdf.lonestar.org> wrote:
>>>
>>>Is there any reason for formatting the drive as FAT32? FAT32 _is_
>>>slow on large hard drives - there is a reason why Microsoft's tools
>>>won't format large FAT32 partitions. I'd suggest retrying the
>>>drive formatted as NTFS.
>>
>> Nope, FAT32 is not much slower. It will not account for
>> this, you would seldom if ever notice a difference in use
>> unless there was a lot of small files moving back and forth
>> over a long period to fragment it, and otherwise only
>> certain types of benchmarks (and even then FAT32 sometimes
>> being faster than NTFS). Besides this, the data rate for
>
>Have you actually looked at the FAT32 data structures? Where is
>the free list or free space bitmap? There isn't one. Finding a
>free sector to write to involves a linear scan of the FAT looking
>for an unallocated cluster. Why have Microsoft introduced exFAT,
>which addresses this issue amongst others, for large volumes where
>the overhead of NTFS is not justified?
>
>FAT's fundamental limitations are fine on small to medium sized
>volumes, but they become significant on larger drives as the FAT
>itself becomes correspondingly larger. NTFS has its own overheads,
>of course, but these are due it it being a robust, modern filesystem
>design. At worst the overhead varies logarithmically with filesystem
>size, not linearly as is the case with any FAT with the same cluster
>size.

You are idealizing instead of seeing things for what they
are, that plenty of people do use FAT32 on large partitions
and don't see any significant performance degradation.
Therefore, and considering the file types aren't hundreds of
thousands of JPGs, rather much larger audio files, it simply
does not matter much, the bottleneck is still drive write
speed not the table. As drives got larger, systems also got
faster. Right now merely reading this your system is doing
something far more demanding that finding a slot in a FAT
table.



>
>> Microsoft's tools won't format larger FAT32 partitions for
>> only one reason - they want to make the industry as
>> proprietary and MS-centric as possible.
>
>Microsoft have claimed several patents in the US regarding FAT32
>(they probably wouldn't be allowed to stand elsewhere). FAT32 is
>already propreitary enough - I'm sure you could find reports of
>the licensing terms on the net if you were so inclined. In any
>case Microsoft have a technically sound justification for this
>decision at http://tinyurl.com/353gcq which also touches on some
>of the other points I have made.

Regardless of what they claimed, they also knew pushing
customers to a proprietary filesystem would make things more
difficult for competitors. We also know all too well how
practically all other semi-popular OS do support FAT32 but
only beginnings of good NTFS support.

No, regardless of whether they feel one is superior or not,
there was no other justification for removing functionality.
Remember, a win9x system can handle larger drives than MS
allowed under FAT32 on 2K/XP, and many have proven they can
do so until it became an issue of other Win9x limitations
with 48bit LBA.

The plain simple truth is anyone who does not need NTFS's
benefits, can format to FAT32 on a large drive and never
know the difference in use. And many do.


>> It's not that I'm against NTFS, it certainly has some
>> advantages like security and > 4GB file support, but
>> performance is not a realistic factor. Then there's the
>> downside, limited to non-existant support of NTFS on a
>> non-*NT* based system.
>
>NTFS support on non-Windows platforms has improved greatly over
>the last three years or so. It isn't really a problem anymore.
>Hell, even NetBSD supports it now, and that tends to stay slightly
>behind the curve in terms of its capabilities.

Yes it has improved, but it's a kludge still, slower in
general than other filesystems on 'nix/etc.



>
>> NTFS fragments. FAT32 can be defragged. Multi-MB music
>> files aren't going to substantially fragment, they are
>> generally written whole and don't fragment at all.
>
>NTFS doesn't fragment to anything like the extent FAT filesystems
>do.

At the same rate, no, but yes it can become fragmented and
reduce performance if we want to think about that.

You continue to try to "like" a filesystem and make theory
instead of seeing fact, that many people do use FAT32 and do
not see the problem the OP is reporting. FAT32 simply isn't
the variable.

If you cared to, you could format a drive to FAT32 and see
this for yourself. Theory about why NTFS is better in your
mind is not an appropriate counter for real evidence of
drives running FAT32 that don't have any significant
problem.

That doesn't mean everyone shouldn't consider the other
"possible" advantages of NTFS, but it does mean it is
wholely inappropriate to try to focus on FAT32 as the cause
of the problem.
 >> Stay informed about: hard drive problem 

On 2007-12-16, kony <spam RemoveThis @spam.com> wrote:
> On Sun, 16 Dec 2007 21:56:16 +0100 (CET), Andrew Smallshaw
><andrews RemoveThis @sdf.lonestar.org> wrote:
>>
>>If it is what I suspect then it is a different problem to what is
>>usually meant by disk fragmentation. If I am correct then it is
>>the FAT itself, not the data on disk, that is causing the problem.
>>A defragmenter may actually make the problem _worse_ since a full
>>defrag will pack all the data at the beginning of the partition.
>>That means that any writes have to pass over all of your exisiting
>>data before some free space can be found.
>
> Hardly.
> Seek time is mere milliseconds. These aren't files that
> would fit in NTFS file table, the very thing you were
> previously claiming a virtue of NTFS is now the thing you
> try to use to condemn FAT32, that NTFS would also have
> packed data and writies further in on the platter. Give it
> up already, it wasn't a FAT32 problem.

It is not the seek time that I am worried about - one of the good
things about FAT32 is that it finally allowed the FAT to be shifted
away from the start of the disk. It is the simple _size_ of the
FAT that causes a problem. A 500Gb FAT32 partition has a FAT over
60Mb in size, even with 32K clusters. Even with the smartest
algorithms out there that is still a hell of a lot of data to sort
through. Defragmentation and packing all files to the start of
the disk _will_ make free space location worse since the search
starts at the beginning where all your data is. Therefore it has
to pass over all of those entries (not the actual data) _first_.

As I said earlier, take a look at the actual data structures if
you don't believe me.

--
Andrew Smallshaw
andrews RemoveThis @sdf.lonestar.org
 >> Stay informed about: hard drive problem 

On Sun, 16 Dec 2007 23:30:48 +0100 (CET), Andrew Smallshaw
<andrews RemoveThis @sdf.lonestar.org> wrote:

>On 2007-12-16, kony <spam RemoveThis @spam.com> wrote:
>> On Sun, 16 Dec 2007 21:56:16 +0100 (CET), Andrew Smallshaw
>><andrews RemoveThis @sdf.lonestar.org> wrote:
>>>
>>>If it is what I suspect then it is a different problem to what is
>>>usually meant by disk fragmentation. If I am correct then it is
>>>the FAT itself, not the data on disk, that is causing the problem.
>>>A defragmenter may actually make the problem _worse_ since a full
>>>defrag will pack all the data at the beginning of the partition.
>>>That means that any writes have to pass over all of your exisiting
>>>data before some free space can be found.
>>
>> Hardly.
>> Seek time is mere milliseconds. These aren't files that
>> would fit in NTFS file table, the very thing you were
>> previously claiming a virtue of NTFS is now the thing you
>> try to use to condemn FAT32, that NTFS would also have
>> packed data and writies further in on the platter. Give it
>> up already, it wasn't a FAT32 problem.
>
>It is not the seek time that I am worried about - one of the good
>things about FAT32 is that it finally allowed the FAT to be shifted
>away from the start of the disk. It is the simple _size_ of the
>FAT that causes a problem. A 500Gb FAT32 partition has a FAT over
>60Mb in size, even with 32K clusters. Even with the smartest
>algorithms out there that is still a hell of a lot of data to sort
>through. Defragmentation and packing all files to the start of
>the disk _will_ make free space location worse since the search
>starts at the beginning where all your data is. Therefore it has
>to pass over all of those entries (not the actual data) _first_.
>
>As I said earlier, take a look at the actual data structures if
>you don't believe me.

Again you are making theories about what may or may not
effect performance without actually trying it to see if it
could actually make anything more than a very small
difference in performance.

The OP couldn't even listen to music, a very low bitrate
either with a file that was already found in the FAT table,
and if MP3, such a very low bitrate that the whole file
might've already been cached into memory.
 >> Stay informed about: hard drive problem