Scandisk and external 320GB USB HD

Franc Zabkar · Jul 17, 2007

I recently purchased a 320GB HD and a USB enclosure. I used Fdisk (in
a Win98SE DOS box) to partition the drive, and I then formatted it via
the GUI. Fdisk reported some strange numbers but in the end it appears
that the drive's full capacity was achieved.

My question is, is it OK to use Scandisk to verify this drive, or will
it corrupt the file system? IIUC, internal drives have issues with
Scandisk but external drives are treated differently in some way. Are
there any other precautions I should be aware of?

BTW, if anyone would like me to do some testing, now is the time. I
suppose I could do my own experiments with Scandisk, but I thought I'd
get some opinions and suggestions before I started.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

MEB · Jul 17, 2007

"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
news:msoo93pvovm5mojhoopa6knemq20tf03ks@4ax.com...
| I recently purchased a 320GB HD and a USB enclosure. I used Fdisk (in
| a Win98SE DOS box) to partition the drive, and I then formatted it via
| the GUI. Fdisk reported some strange numbers but in the end it appears
| that the drive's full capacity was achieved.
|
| My question is, is it OK to use Scandisk to verify this drive, or will
| it corrupt the file system? IIUC, internal drives have issues with
| Scandisk but external drives are treated differently in some way. Are
| there any other precautions I should be aware of?
|
| BTW, if anyone would like me to do some testing, now is the time. I
| suppose I could do my own experiments with Scandisk, but I thought I'd
| get some opinions and suggestions before I started.
|
| - Franc Zabkar
| --
| Please remove one 'i' from my address when replying by email.

Of course we would, how about that 4kb / several hundred million files on a
single fat32 partition test that 98 Guy was supposed to do... [though that
was SATA so ....] but really, what kind of testing do YOU want to do... {and
did you buy me one too???]

Umm, partition size(s)??

Ingeborg · Jul 17, 2007

Franc Zabkar wrote:

> I recently purchased a 320GB HD and a USB enclosure. I used Fdisk (in
> a Win98SE DOS box) to partition the drive, and I then formatted it via
> the GUI. Fdisk reported some strange numbers but in the end it appears
> that the drive's full capacity was achieved.
>
> My question is, is it OK to use Scandisk to verify this drive, or will
> it corrupt the file system? IIUC, internal drives have issues with
> Scandisk but external drives are treated differently in some way. Are
> there any other precautions I should be aware of?
>

Scandisk will corrupt your filesystem.

There is a difference on disklevel between internal and external disks
(internal disks use LBA28 addressing, which give you a maximum of 2^28
sectors, which is 128 GiB), but 'the scandisk bug' is a memory limitation,
and has nothing to do with sector addresses.
<http://support.microsoft.com/kb/184006>

WinME's scandisk doesn't have this limitation.

Curt Christianson · Jul 17, 2007

Hi Franc,

Long time since I've spoken with you. I believe MS-MVP Chris Quirke has
written about Scandisk "breaking" large drives. I'll try to find the info.,
in the meantime maybe he'll see this and jump in.

Thanks for all the expertise you provide.

--
HTH,
Curt

Windows Support Center
www.aumha.org
Practically Nerded,...
http://dundats.mvps.org/Index.htm

"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
news:msoo93pvovm5mojhoopa6knemq20tf03ks@4ax.com...
|I recently purchased a 320GB HD and a USB enclosure. I used Fdisk (in
| a Win98SE DOS box) to partition the drive, and I then formatted it via
| the GUI. Fdisk reported some strange numbers but in the end it appears
| that the drive's full capacity was achieved.
|
| My question is, is it OK to use Scandisk to verify this drive, or will
| it corrupt the file system? IIUC, internal drives have issues with
| Scandisk but external drives are treated differently in some way. Are
| there any other precautions I should be aware of?
|
| BTW, if anyone would like me to do some testing, now is the time. I
| suppose I could do my own experiments with Scandisk, but I thought I'd
| get some opinions and suggestions before I started.
|
| - Franc Zabkar
| --
| Please remove one 'i' from my address when replying by email.

98 Guy · Jul 17, 2007

Ingeborg wrote:

> > I recently purchased a 320GB HD and a USB enclosure.
> > My question is, is it OK to use Scandisk to verify this drive,
>
> Scandisk will corrupt your filesystem.

Do you/we know if USB drives are controlled via ESDI_506.PDR?

What do they say on MSFN about that?

Ingeborg · Jul 17, 2007

98 Guy wrote:

> Ingeborg wrote:
>
>> > I recently purchased a 320GB HD and a USB enclosure.
>> > My question is, is it OK to use Scandisk to verify this drive,
>>
>> Scandisk will corrupt your filesystem.
>
> Do you/we know if USB drives are controlled via ESDI_506.PDR?
>

Yes I know, they don't. But this has knothing to do with the scandisk bug.

Franc Zabkar · Jul 18, 2007

On Tue, 17 Jul 2007 03:01:24 -0400, "MEB" <meb@not here@hotmail.com>
put finger to keyboard and composed:

>
>"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
>news:msoo93pvovm5mojhoopa6knemq20tf03ks@4ax.com...
>| I recently purchased a 320GB HD and a USB enclosure. I used Fdisk (in
>| a Win98SE DOS box) to partition the drive, and I then formatted it via
>| the GUI. Fdisk reported some strange numbers but in the end it appears
>| that the drive's full capacity was achieved.
>|
>| My question is, is it OK to use Scandisk to verify this drive, or will
>| it corrupt the file system? IIUC, internal drives have issues with
>| Scandisk but external drives are treated differently in some way. Are
>| there any other precautions I should be aware of?
>|
>| BTW, if anyone would like me to do some testing, now is the time. I
>| suppose I could do my own experiments with Scandisk, but I thought I'd
>| get some opinions and suggestions before I started.
>|
>| - Franc Zabkar
>| --
>| Please remove one 'i' from my address when replying by email.
>
>
> Of course we would, how about that 4kb / several hundred million files on a
>single fat32 partition test that 98 Guy was supposed to do... [though that
>was SATA so ....] but really, what kind of testing do YOU want to do... {and
>did you buy me one too???]
>
> Umm, partition size(s)??

My ultimate use for this drive will be as a DivX video jukebox
attached to the USB input of my DVD player. The chosen configuration
will need to be compatible with both the player and my Win98SE boxes.
I have yet to find out whether the player can handle a drive with more
than one partition, but so far it seems to be OK with a single 320GB
partition. My one major concern is that my data may be vulnerable to
corruption if I accidentally run some incompatible disc utility.

The main reason for my post was to offer this opportunity to
contribute to the body of test results posted by 98Guy in February
2007.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

Franc Zabkar · Jul 18, 2007

On Tue, 17 Jul 2007 03:46:35 -0700, Ingeborg <a@b.invalid> put finger
to keyboard and composed:

>Franc Zabkar wrote:
>
>> I recently purchased a 320GB HD and a USB enclosure. I used Fdisk (in
>> a Win98SE DOS box) to partition the drive, and I then formatted it via
>> the GUI. Fdisk reported some strange numbers but in the end it appears
>> that the drive's full capacity was achieved.
>>
>> My question is, is it OK to use Scandisk to verify this drive, or will
>> it corrupt the file system? IIUC, internal drives have issues with
>> Scandisk but external drives are treated differently in some way. Are
>> there any other precautions I should be aware of?
>>
>
>Scandisk will corrupt your filesystem.
>
>There is a difference on disklevel between internal and external disks
>(internal disks use LBA28 addressing, which give you a maximum of 2^28
>sectors, which is 128 GiB), but 'the scandisk bug' is a memory limitation,
>and has nothing to do with sector addresses.
><http://support.microsoft.com/kb/184006>
>
>WinME's scandisk doesn't have this limitation.

Thanks. I remember 98Guy's extensive posts on this subject but I
confess that I didn't absorb too much at that time. However, I am now
in a position to follow them up in a practical sense.

IIUC, Win98SE's disk utilities (eg Scandskw and Defrag) will be OK
with my drive provided that I split it into partitions which are
smaller than 127GB, to accommodate a FAT size limitation. Otherwise,
for a single 320GB partition, I should use WinME's versions of
scandskw.exe, dskmaint.dll and defrag.exe, none of which have this
limitation. Alternatively, I should remember not to run the Win98SE
versions of these utilities. As for the LBA28 limitation, my drive is
not affected because it does not rely on the ESDI_506.PDR driver (I'm
using the Maximus Decim NUSB drivers).

Have I got this right?

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

Ingeborg · Jul 18, 2007

Franc Zabkar wrote:

> On Tue, 17 Jul 2007 03:46:35 -0700, Ingeborg <a@b.invalid> put finger
> to keyboard and composed:
>
>>Franc Zabkar wrote:
>>
>>> I recently purchased a 320GB HD and a USB enclosure. I used Fdisk
>>> (in a Win98SE DOS box) to partition the drive, and I then formatted
>>> it via the GUI. Fdisk reported some strange numbers but in the end
>>> it appears that the drive's full capacity was achieved.
>>>
>>> My question is, is it OK to use Scandisk to verify this drive, or
>>> will it corrupt the file system? IIUC, internal drives have issues
>>> with Scandisk but external drives are treated differently in some
>>> way. Are there any other precautions I should be aware of?
>>>
>>
>>Scandisk will corrupt your filesystem.
>>
>>There is a difference on disklevel between internal and external disks
>>(internal disks use LBA28 addressing, which give you a maximum of 2^28
>>sectors, which is 128 GiB), but 'the scandisk bug' is a memory
>>limitation, and has nothing to do with sector addresses.
>><http://support.microsoft.com/kb/184006>
>>
>>WinME's scandisk doesn't have this limitation.
>
> Thanks. I remember 98Guy's extensive posts on this subject but I
> confess that I didn't absorb too much at that time. However, I am now
> in a position to follow them up in a practical sense.
>
> IIUC, Win98SE's disk utilities (eg Scandskw and Defrag) will be OK
> with my drive provided that I split it into partitions which are
> smaller than 127GB, to accommodate a FAT size limitation. Otherwise,
> for a single 320GB partition, I should use WinME's versions of
> scandskw.exe, dskmaint.dll and defrag.exe, none of which have this
> limitation. Alternatively, I should remember not to run the Win98SE
> versions of these utilities. As for the LBA28 limitation, my drive is
> not affected because it does not rely on the ESDI_506.PDR driver (I'm
> using the Maximus Decim NUSB drivers).
>
> Have I got this right?
>

Yes.

MEB · Jul 18, 2007

"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
news:c9jq93pir9u5opug01a3hnp07b3tntiha7@4ax.com...
| On Tue, 17 Jul 2007 03:01:24 -0400, "MEB" <meb@not here@hotmail.com>
| put finger to keyboard and composed:
|
| >
| >"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
| >news:msoo93pvovm5mojhoopa6knemq20tf03ks@4ax.com...
| >| I recently purchased a 320GB HD and a USB enclosure. I used Fdisk (in
| >| a Win98SE DOS box) to partition the drive, and I then formatted it via
| >| the GUI. Fdisk reported some strange numbers but in the end it appears
| >| that the drive's full capacity was achieved.
| >|
| >| My question is, is it OK to use Scandisk to verify this drive, or will
| >| it corrupt the file system? IIUC, internal drives have issues with
| >| Scandisk but external drives are treated differently in some way. Are
| >| there any other precautions I should be aware of?
| >|
| >| BTW, if anyone would like me to do some testing, now is the time. I
| >| suppose I could do my own experiments with Scandisk, but I thought I'd
| >| get some opinions and suggestions before I started.
| >|
| >| - Franc Zabkar
| >| --
| >| Please remove one 'i' from my address when replying by email.
| >
| >
| > Of course we would, how about that 4kb / several hundred million files
on a
| >single fat32 partition test that 98 Guy was supposed to do... [though
that
| >was SATA so ....] but really, what kind of testing do YOU want to do...
{and
| >did you buy me one too???]
| >
| > Umm, partition size(s)??
|
| My ultimate use for this drive will be as a DivX video jukebox
| attached to the USB input of my DVD player. The chosen configuration
| will need to be compatible with both the player and my Win98SE boxes.
| I have yet to find out whether the player can handle a drive with more
| than one partition, but so far it seems to be OK with a single 320GB
| partition. My one major concern is that my data may be vulnerable to
| corruption if I accidentally run some incompatible disc utility.
|
| The main reason for my post was to offer this opportunity to
| contribute to the body of test results posted by 98Guy in February
| 2007.
|
| - Franc Zabkar
| --
| Please remove one 'i' from my address when replying by email.

That was understood, of course most of that information was already
reviewable on MSFN, Axcel216, and even Microsoft [and several other sites
and articles which appear in those threads and elsewhere]... however, as
noted in these discussions, the esdi_506.pdr is not used [as in 98 Guy's
post and now here] so potentially the hundreds of millions of directory and
files test would likely STILL be viable under these conditions [though with
the notation of his RAID and SATA, and your USB].
The test is to stress the fat tables, parsing, memory loads, etc. to which
98 Guy alleged / discussions were held / and resolution/proof was not
actually presented: e.g., never tested/stressed with his massively large
files and relatively few entries... though that would still leave the OS on
the same large partition and any issues which might be related thereto... so
if you wish, for clarification sake, the test would still appear to be
viable and useful...

98 Guy · Jul 18, 2007

MEB wrote:

> ... however, as noted in these discussions, the esdi_506.pdr is
> not used [as in 98 Guy's post and now here] so potentially the
> hundreds of millions of directory and files test would likely
> STILL be viable under these conditions.

> The test is to stress the fat tables, parsing, memory loads,
> etc. to which 98 Guy alleged / discussions were held / and
> resolution/proof was not actually presented: e.g., never tested
> /stressed with his massively large files and relatively few entries

Would you care to explain the nature or theory behind a file-count
limitation that win-98 (or FAT-32?) *may* be vulnerable to - or
reference someone else's intimation of such a limitation?

And could you also explain how such a limitation couldn't be
experienced given a 127 gb volume?

MEB · Jul 19, 2007

"98 Guy" <98@Guy.com> wrote in message news:469E8E17.FB6FF64F@Guy.com...
| MEB wrote:
|
| > ... however, as noted in these discussions, the esdi_506.pdr is
| > not used [as in 98 Guy's post and now here] so potentially the
| > hundreds of millions of directory and files test would likely
| > STILL be viable under these conditions.
|
| > The test is to stress the fat tables, parsing, memory loads,
| > etc. to which 98 Guy alleged / discussions were held / and
| > resolution/proof was not actually presented: e.g., never tested
| > /stressed with his massively large files and relatively few entries
|
| Would you care to explain the nature or theory behind a file-count
| limitation that win-98 (or FAT-32?) *may* be vulnerable to - or
| reference someone else's intimation of such a limitation?
|
| And could you also explain how such a limitation couldn't be
| experienced given a 127 gb volume?

WHAT?? How many people have to explain this to you... you had what,, four or
six in the last two discussions.. you were shown the documentation in every
one of the discussions and other aspects, and still, YOU NEVER DID THE TEST
....
So do it, prove it or disprove it.. wonderful discussions, lots of good
information, but proof the OS and fat can handle it on a daily basis... what
part of that don't you get?

--
MEB
http://peoplescounsel.orgfree.com
________

MEB · Jul 19, 2007

"98 Guy" <98@Guy.com> wrote in message news:469E8E17.FB6FF64F@Guy.com...
| MEB wrote:
|
| > ... however, as noted in these discussions, the esdi_506.pdr is
| > not used [as in 98 Guy's post and now here] so potentially the
| > hundreds of millions of directory and files test would likely
| > STILL be viable under these conditions.
|
| > The test is to stress the fat tables, parsing, memory loads,
| > etc. to which 98 Guy alleged / discussions were held / and
| > resolution/proof was not actually presented: e.g., never tested
| > /stressed with his massively large files and relatively few entries
|
| Would you care to explain the nature or theory behind a file-count
| limitation that win-98 (or FAT-32?) *may* be vulnerable to - or
| reference someone else's intimation of such a limitation?
|
| And could you also explain how such a limitation couldn't be
| experienced given a 127 gb volume?

WHAT?? How many people have to explain this to you... you had what,, four or
six in the last two discussions.. you were shown the documentation in every
one of the discussions and other aspects, and still,
YOU NEVER DID THE TEST to prove it one way or the other, that's all the
required..
You DID NOT test it on the 127 gig.... at least not here in this group..
....
So do it, prove it or disprove it.. wonderful discussions, lots of good
information, but proof the OS and fat can handle it on a daily basis... what
part of that don't you get?

--
MEB
http://peoplescounsel.orgfree.com
________

98 Guy · Jul 19, 2007

MEB wrote:

> | Would you care to explain the nature or theory behind a file-count
> | limitation that win-98 (or FAT-32?) *may* be vulnerable to - or
> | reference someone else's intimation of such a limitation?
> |
> | And could you also explain how such a limitation couldn't be
> | experienced given a 127 gb volume?
>
> WHAT?? How many people have to explain this to you...

Again MEB fails to answer a simple question, but flies off the handle
and brushes the question away. Why even bother to respond?

> you had what,, four or six in the last two discussions..
> you were shown the documentation in every one of the
> discussions and other aspects,

So dig up the relavent material from those posts that answers my
question.

> and still, YOU NEVER DID THE TEST

That is not the point. Don't change the subject.

Where in those posts was there some supporting backup for this
supposed file-count limitation?

And listen to this:

Since I formatted a 160 gb drive using 4kb clusters, resulting in 40
million clusters, that would theoretically mean a maximum of 40
million files (you can't store more than 1 file in a single cluster).
Your claim that it's possible to have *hundreds* of millions of files
is baloney.

The file-count limitation (if it exists) is related to the number of
clusters on a volume. Given a 500 gb volume formatted with 32kb
cluster size, that would result in around 15 million clusters. Just
when is this supposed file-count limitation supposed to kick in?

> So do it, prove it or disprove it..

There is good background and history regarding the 127 gb volume-size
limitation caused by ESDI_506.PDR.

All I'm asking for is some background in the origination of where this
file-count limitation theory comes from. Some history. If you can't
provide it, don't froth at the mouth and obfuscate the question.

MEB · Jul 19, 2007

Look, that test has ALWAYS been the point, without the true results all your
postings mean *squat*.

98 Guy · Jul 19, 2007

MEB wrote:

> Look, that test has ALWAYS been the point

I'm not arguing for or against the existance of the file-count
limitation.

> without the true results all your postings mean *squat*.

You really don't get it, do you?

Whether or not I or anyone else performs *the test* is one thing.

I'm asking you (or anyone else reading this) to point to where this
file-count limitation was first propositioned, or by whom. Such
information would hopefully include a theory as to why win-98 might
have such a limitation, perhaps descibing the underlying mechanism
causing the limitation.

I don't think that's a foolish or irrational or pointless question,
and it can be asked in conjunction with requests to perform *the
test*. And it's a question that you should be able to answer, given
that you seem to know so much about this issue. Nobody else here is
stepping up and offering an answer.

What I have found is this:

- The claimed max number of files per directory is 65535.

http://www.microsoft.com/technet/technetmag/issues/2006/07/WindowsConfidential/

- Max number of files possible using FAT-32: 268,435,437

http://en.wikipedia.org/wiki/File_Allocation_Table#Design

The number 268,435,437 (max number of files possible under FAT32) is
equal to 16k x 16k (ie, 16k directories, each with 16k files) or 4k x
64k (ie - 4k directories, each with 64k files).

A check of one of my win-98 systems shows 4,870 folders and 118k
files, so clearly FAT-32 can have more than 4k directories, and win-98
can handle at least that many directories as well.

Again, if there is a source, a document, a person, that claims (or
theorizes) that win-98 is practically limited to some particular
number of directories or files per volume, then please post that
information.

MEB · Jul 19, 2007

Look dude, I'd plonk your stupid stuff, but then you would have free range
here...

These are the EXACT questions you asked and were answered repeatedly in ALL
your discussions, I'm NOT doing the same stuff with you over and over
again,, I don't think anyone will...

But I will continue to point out the lack of any supporting evidence via
testing to support your claims... and the other crap you post here which is
counter to security or OS stability... count on it

"98 Guy" <98@Guy.com> wrote in message news:469FDC57.55ABEE46@Guy.com...
| MEB wrote:
|
| > Look, that test has ALWAYS been the point
|
| I'm not arguing for or against the existance of the file-count
| limitation.
|
| > without the true results all your postings mean *squat*.
|
| You really don't get it, do you?
|
| Whether or not I or anyone else performs *the test* is one thing.
|
| I'm asking you (or anyone else reading this) to point to where this
| file-count limitation was first propositioned, or by whom. Such
| information would hopefully include a theory as to why win-98 might
| have such a limitation, perhaps descibing the underlying mechanism
| causing the limitation.
|
| I don't think that's a foolish or irrational or pointless question,
| and it can be asked in conjunction with requests to perform *the
| test*. And it's a question that you should be able to answer, given
| that you seem to know so much about this issue. Nobody else here is
| stepping up and offering an answer.
|
| What I have found is this:
|
| - The claimed max number of files per directory is 65535.
|
|
http://www.microsoft.com/technet/technetmag/issues/2006/07/WindowsConfidential/
|
| - Max number of files possible using FAT-32: 268,435,437
|
| http://en.wikipedia.org/wiki/File_Allocation_Table#Design
|
| The number 268,435,437 (max number of files possible under FAT32) is
| equal to 16k x 16k (ie, 16k directories, each with 16k files) or 4k x
| 64k (ie - 4k directories, each with 64k files).
|
| A check of one of my win-98 systems shows 4,870 folders and 118k
| files, so clearly FAT-32 can have more than 4k directories, and win-98
| can handle at least that many directories as well.
|
| Again, if there is a source, a document, a person, that claims (or
| theorizes) that win-98 is practically limited to some particular
| number of directories or files per volume, then please post that
| information.

98 Guy · Jul 19, 2007

MEB wrote:

> These are the EXACT questions you asked and were answered
> repeatedly in ALL your discussions,

I'm asking what is the source for the speculation that win-98 has a
built-in file-count limitation. You never answered that question. If
the answer exists in any thread, in any public forum, then please
quote it.

> I'm NOT doing the same stuff with you over and over again,,

You never went over it A FIRST TIME.

> But I will continue to point out the lack of any supporting
> evidence via testing to support your claims...

You have a reading comprehension problem.

I have not stated anywhere that win-98 is, or is not, limited in the
number of files it can handle on a FAT-32 volume.

I continue to ask you where did such a notion about this limitation
come from? What is the source of such an idea? Is there any
supporting facts for such a limitation?

If you can't answer that question, then just say so.

All you have to say is this:

"I don't know who got the idea that win-98 might have
an upper limit to the number of files on a volume that
it can handle reliably or efficiently."

"I don't know when such an idea was first raised in
a public forum. I don't know if any technical reasons
were speculated as to why there might be such a
limitation."

"All I know is that we are talking about it now, in
this forum, and a methodology was published some
time ago to determine if such a limitation exists."

That's all you have to say. I've already typed it in for you. Just
repeat it in your next reply.

Franc Zabkar · Jul 20, 2007

On Thu, 19 Jul 2007 17:49:11 -0400, 98 Guy <98@Guy.com> put finger to
keyboard and composed:

>MEB wrote:
>
>> Look, that test has ALWAYS been the point
>
>I'm not arguing for or against the existance of the file-count
>limitation.
>
>> without the true results all your postings mean *squat*.
>
>You really don't get it, do you?
>
>Whether or not I or anyone else performs *the test* is one thing.
>
>I'm asking you (or anyone else reading this) to point to where this
>file-count limitation was first propositioned, or by whom. Such
>information would hopefully include a theory as to why win-98 might
>have such a limitation, perhaps descibing the underlying mechanism
>causing the limitation.
>
>I don't think that's a foolish or irrational or pointless question,
>and it can be asked in conjunction with requests to perform *the
>test*. And it's a question that you should be able to answer, given
>that you seem to know so much about this issue. Nobody else here is
>stepping up and offering an answer.
>
>What I have found is this:
>
>- The claimed max number of files per directory is 65535.
>
>http://www.microsoft.com/technet/technetmag/issues/2006/07/WindowsConfidential/
>
>- Max number of files possible using FAT-32: 268,435,437
>
>http://en.wikipedia.org/wiki/File_Allocation_Table#Design
>
>The number 268,435,437 (max number of files possible under FAT32) is
>equal to 16k x 16k (ie, 16k directories, each with 16k files) or 4k x
>64k (ie - 4k directories, each with 64k files).
>
>A check of one of my win-98 systems shows 4,870 folders and 118k
>files, so clearly FAT-32 can have more than 4k directories, and win-98
>can handle at least that many directories as well.
>
>Again, if there is a source, a document, a person, that claims (or
>theorizes) that win-98 is practically limited to some particular
>number of directories or files per volume, then please post that
>information.

I just posted the following question to alt.msdos.batch:

==================================================================
I'm looking for a batch routine to create a tree of 65536 directories
with 16 at the top level, 256 subdirectories at the next level, 4096
at the next, and finally 65536. The reason is that I wish to test the
hypothesis that there is a limit to the number of directories in a
FAT32 file system.
==================================================================

One person responded with this:

==================================================================
Here you can see Windows properties screenshots showing FAT32
filesystem and the directory count [1,118,480]:

http://guysalias.batcave.net/images/a-lot-of-directories.jpg

FAT32 does not have a limit to the number of files or directories
except that imposed by the size of the drive(number of clusters).
==================================================================

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

MEB · Jul 20, 2007

"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
news:qds0a3lvjmrahn3p41lh7npmqr9ec4i3v2@4ax.com...
| On Thu, 19 Jul 2007 17:49:11 -0400, 98 Guy <98@Guy.com> put finger to
| keyboard and composed:
|
| >MEB wrote:
| >
| >> Look, that test has ALWAYS been the point
| >
| >I'm not arguing for or against the existance of the file-count
| >limitation.
| >
| >> without the true results all your postings mean *squat*.
| >
| >You really don't get it, do you?
| >
| >Whether or not I or anyone else performs *the test* is one thing.
| >
| >I'm asking you (or anyone else reading this) to point to where this
| >file-count limitation was first propositioned, or by whom. Such
| >information would hopefully include a theory as to why win-98 might
| >have such a limitation, perhaps descibing the underlying mechanism
| >causing the limitation.
| >
| >I don't think that's a foolish or irrational or pointless question,
| >and it can be asked in conjunction with requests to perform *the
| >test*. And it's a question that you should be able to answer, given
| >that you seem to know so much about this issue. Nobody else here is
| >stepping up and offering an answer.
| >
| >What I have found is this:
| >
| >- The claimed max number of files per directory is 65535.
| >
|
>http://www.microsoft.com/technet/technetmag/issues/2006/07/WindowsConfident
ial/
| >
| >- Max number of files possible using FAT-32: 268,435,437
| >
| >http://en.wikipedia.org/wiki/File_Allocation_Table#Design
| >
| >The number 268,435,437 (max number of files possible under FAT32) is
| >equal to 16k x 16k (ie, 16k directories, each with 16k files) or 4k x
| >64k (ie - 4k directories, each with 64k files).
| >
| >A check of one of my win-98 systems shows 4,870 folders and 118k
| >files, so clearly FAT-32 can have more than 4k directories, and win-98
| >can handle at least that many directories as well.
| >
| >Again, if there is a source, a document, a person, that claims (or
| >theorizes) that win-98 is practically limited to some particular
| >number of directories or files per volume, then please post that
| >information.
|
| I just posted the following question to alt.msdos.batch:
|
| ==================================================================
| I'm looking for a batch routine to create a tree of 65536 directories
| with 16 at the top level, 256 subdirectories at the next level, 4096
| at the next, and finally 65536. The reason is that I wish to test the
| hypothesis that there is a limit to the number of directories in a
| FAT32 file system.
| ==================================================================
|
| One person responded with this:
|
| ==================================================================
| Here you can see Windows properties screenshots showing FAT32
| filesystem and the directory count [1,118,480]:
|
| http://guysalias.batcave.net/images/a-lot-of-directories.jpg
|
| FAT32 does not have a limit to the number of files or directories
| except that imposed by the size of the drive(number of clusters).
| ==================================================================
|
| - Franc Zabkar
| --
| Please remove one 'i' from my address when replying by email.

Excellent, part of an answer with test results. Though distinctly limited.

It shows ONLY a 4.87 gig drive/partition and just 1,118,480 folders. It
also fails to show exactly which OS was used in testing.
There is the inclusion of the hardware tag and sharing [indicating
networking], bring some special aspects which may have been involved. SO is
or was this 98 or is this an NT aspect which handles the Fat [and drives] in
a different fashion?

A single partition on a 320 [or the 127 per the other discussion] gig drive
would potentially have how many using that same test?

SO did you find any test results:
1. showing usage of 98 showing any physical manipulations and,
2. usage on a DAILY BASIS, e.g. the normal cut/paste, delete, move,
yada-yada, of the potential hundreds of millions of files and directories on
drives and
3. which exceed the 128 gig purported limitations, and
4. partitioned as a single drive running 98 exclusively, and
5. that might be found during usage over extended periods, and
6. in a situation of such daily usage and/or file serving/storage stature,
and
7. including of course the non-standard 4kb aspect which was purported as
non-issue?

--
MEB
http://peoplescounsel.orgfree.com
________

Scandisk and external 320GB USB HD

Franc Zabkar

MEB

Ingeborg

Curt Christianson

98 Guy

Ingeborg

Franc Zabkar

Franc Zabkar

Ingeborg

MEB

98 Guy

MEB

MEB

98 Guy

MEB

98 Guy

MEB

98 Guy

Franc Zabkar

MEB

Similar threads