Partitions disappeared

[Annika K]

I am runing Windows 98 on a 200 Gig HD. There are 3 partitions on the HD.
One is about 10 gigs and holds the operating system. Yesterday the other
two partitions disappered while windows was running. So far I have rebooted
a couple of times (before i realized what was going on) and attempted to fix
the mbr with "FDISK /MBR" which did not fix the problem. I have run
FindPart and will include the ouput from that below. But after this point I
am out of my league. Is the output from FindPart enough to restore the
partitions? I am not concerned with the data on the OS partition. I am
fine with having to reinstall the OS. It's the data on the other 2 drives
that I am interested in. Can someone please guide me through this?

Thank you very much.

FindPart Output....................

Findpart, version 4.42.
Copyright Svend Olaf Mikkelsen, 1999-2004.

OS: DOS 7.10

Disk: 1 Cylinders: 24792 Heads: 255 Sectors: 63 MB: 194474

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 - 0C 63 19534977 9538 0 1 1 1215 254 63 B OK
8511 1 0B 63261554202127712 8511 1 1 24791 254 63 OK OK

------FAT CHS -Size Cl --Root -Good -Rep. Maybe --Bad YY-MM-DD DataMB
0 1 33 9530 8 2 9530 0 0 0 05-03-07 454
8511 1 33 31921 32 2 31921 0 0 0 05-02-21 75808
13317 0 33 Second FAT not found.

Partitions according to partition tables on first harddisk:

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 1*0C 63 19534977 9538 0 1 1 1215 254 63 OK OK
0 2 0F 19535040378748440184935 1216 0 1 24791 254 63 OK

No signature CHS: 1216 0 1

 

[Lil' Dave]

I am runing Windows 98 on a 200 Gig HD. There are 3 partitions on the HD.
One is about 10 gigs and holds the operating system. Yesterday the other
two partitions disappered while windows was running. So far I have rebooted
a couple of times (before i realized what was going on) and attempted to fix
the mbr with "FDISK /MBR" which did not fix the problem. I have run
FindPart and will include the ouput from that below. But after this point I
am out of my league. Is the output from FindPart enough to restore the
partitions? I am not concerned with the data on the OS partition. I am
fine with having to reinstall the OS. It's the data on the other 2 drives
that I am interested in. Can someone please guide me through this?

Thank you very much.

FindPart Output....................

Findpart, version 4.42.
Copyright Svend Olaf Mikkelsen, 1999-2004.

OS: DOS 7.10

Disk: 1 Cylinders: 24792 Heads: 255 Sectors: 63 MB: 194474

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 - 0C 63 19534977 9538 0 1 1 1215 254 63 B OK
8511 1 0B 63261554202127712 8511 1 1 24791 254 63 OK OK

------FAT CHS -Size Cl --Root -Good -Rep. Maybe --Bad YY-MM-DD DataMB
0 1 33 9530 8 2 9530 0 0 0 05-03-07 454
8511 1 33 31921 32 2 31921 0 0 0 05-02-21 75808
13317 0 33 Second FAT not found.

Partitions according to partition tables on first harddisk:

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 1*0C 63 19534977 9538 0 1 1 1215 254 63 OK OK
0 2 0F 19535040378748440184935 1216 0 1 24791 254 63 OK

No signature CHS: 1216 0 1

You did not indicate the size of the missing 2 partitions. Your mistake was
running fdisk /mbr as win98's fdisk only "understands" up to 64GB size
partitions. After that, you get cylinder wrap. An updated version of fdisk
from MS does up to 128GB.

 

[Folkert Rienstra]

You did not indicate the size of the missing 2 partitions. Your mistake was
running fdisk /mbr as win98's fdisk only "understands" up to 64GB size
partitions. After that, you get cylinder wrap. An updated version of fdisk
from MS does up to 128GB.

Wotanidiot.

 

[Annika K]

Is it too late to use the updated fdisk?

You did not indicate the size of the missing 2 partitions. Your mistake
was
running fdisk /mbr as win98's fdisk only "understands" up to 64GB size
partitions. After that, you get cylinder wrap. An updated version of
fdisk
from MS does up to 128GB.

 

[Eric Gisin]

You did not indicate the size of the missing 2 partitions. Your mistake was
running fdisk /mbr as win98's fdisk only "understands" up to 64GB size
partitions. After that, you get cylinder wrap. An updated version of fdisk
from MS does up to 128GB.

Fdisk /mbr never fixes such problems. The original version works, it just
reports the wrong drive size.

If the OP has an older machine, the BIOS does not support over 135GB. You will
certainly get corruption with a 200GB disk and Win98.

 

[Annika K]

It's not an older machine..at least I don’t think so. The BIOS recognized
the hard drive correctly. How old is older? At any rate, is there any hope
for me or is all lost since I ran FDISK /mbr? If there is hope can someone
walk me through this or point to a site that tells me precisely how to
recover the partitions? I'm in over my head on this one.

Thanks very much

 

[Folkert Rienstra]

I am runing Windows 98 on a 200 Gig HD. There are 3 partitions on the HD.
One is about 10 gigs and holds the operating system. Yesterday the other
two partitions disappered while windows was running. So far I have rebooted
a couple of times (before i realized what was going on) and attempted to fix
the mbr with "FDISK /MBR" which did not fix the problem.

That only fixes (very limited) bootproblems.

I have run FindPart and will include the ouput from that below.
But after this point I am out of my league.

Is the output from FindPart enough to restore the partitions?

It can usually be used to serve as input to a partition table editor.

I am not concerned with the data on the OS partition. I am
fine with having to reinstall the OS. It's the data on the other 2 drives
that I am interested in. Can someone please guide me through this?

Thank you very much.

FindPart Output....................

Findpart, version 4.42.
Copyright Svend Olaf Mikkelsen, 1999-2004.

OS: DOS 7.10

Disk: 1 Cylinders: 24792 Heads: 255 Sectors: 63 MB: 194474

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS

0 - 0C 63 19534977 9538 0 1 1 1215 254 63 B OK

The bootrecord of the primary partition, drive C:

8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

The bootrecord of the second logical, drive E:
The bootrecord of the first logical, (drive D at cyl. 1216 is missing

------FAT CHS -Size Cl --Root -Good -Rep. Maybe --Bad YY-MM-DD DataMB
0 1 33 9530 8 2 9530 0 0 0 05-03-07 454

FAT of the primary partition, drive C:

8511 1 33 31921 32 2 31921 0 0 0 05-02-21 75808

FATs of 2nd logical again.
The FATs of 1st logical at cyl 1216 are missing.

13317 0 33 Second FAT not found.

Partitions according to partition tables on first harddisk:

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS

0 1*0C 63 19534977 9538 0 1 1 1215 254 63 OK OK

MBR, primary partition, drive C:

0 2 0F 19535040 378748440 184935 1216 0 1 24791 254 63 OK

Extended partition, holder of logicals D: and E:
The EMBR at cyl. 1216 that the Extended partition points to and it's Logical
and the pointer to the second EMBR at cyl. 8511 and its logical are missing.

The EMBR at cyl. 8511 itself may be OK.

No signature (at) CHS: 1216 0 1

May be the reason why the 1st EMBR isn't listed.
Inconclusive since there is no such comment on the second EMBR and it isn't
listed either.

Recreating the first EMBR may return access to drive E:
Drive D: is lost.

 

[Joep]

I am runing Windows 98 on a 200 Gig HD. There are 3 partitions on the HD.
One is about 10 gigs and holds the operating system. Yesterday the other
two partitions disappered while windows was running. So far I have rebooted
a couple of times (before i realized what was going on) and attempted to fix
the mbr with "FDISK /MBR" which did not fix the problem. I have run
FindPart and will include the ouput from that below. But after this point I
am out of my league. Is the output from FindPart enough to restore the
partitions? I am not concerned with the data on the OS partition. I am
fine with having to reinstall the OS. It's the data on the other 2 drives
that I am interested in. Can someone please guide me through this?

Hi,

If Svend doesn't reply you may want to try DiskPatch from
www.diydatarecovery.nl

 

[Annika K]

Thank you, Folkert. Do I use ptedit to recreate the EMBR? I read
instructions on this page
http://www.warpdoctor.org/walter/articles/2000/aa061400a.html
that say to change the type column from 0F to 05. Is this what i need to
do?

 

[Folkert Rienstra]

Thank you, Folkert. Do I use ptedit to recreate the EMBR?

That is what I prefer.

I read instructions on this page
http://www.warpdoctor.org/walter/articles/2000/aa061400a.html
that say to change the type column from 0F to 05.

Is this what i need to do?

Possibly. I have no idea what that is about without further examination.

Lines copied from other post, for reference:

Disk: 1 Cylinders: 24792 Heads: 255 Sectors: 63 MB: 194474

Partitions according to partition bootrecords on first harddisk:
--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 - 0C 63 19534977 9538 0 1 1 1215 254 63 B OK
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

Partitions according to partition tables on first harddisk:
--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 1*0C 63 19534977 9538 0 1 1 1215 254 63 OK OK
0 2 0F 19535040 378748440 184935 1216 0 1 24791 254 63 OK

Punch the EMBR (extended partition MBR) button in the opening screen and
put most of the info of this line:

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

into line 2 of the new tables presented, like so:

type: 0F starting*: 8511 0 1, ending*: 24791 254 63, sectors before: xxxxxxxxx, sectors: 261554202,
where xxxxxxxxx = (8510+1)*(254+1)*63 = 136729215 (Cylinders and heads are 0-based)
* Starting and ending CHS will not be accepted for cylinders over 1023, you can fill
in the maximum CHS that is allowed for them: 1023 254 63, i.e. large drive placeholders.

If all is well then ptedit should now let you go to the next EMBR for the just defined
(2nd) extended partition and show you the logical (E in it, as represented by:

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

and you should be able to open the bootrecord.

Next:
You can combine the info of these 2 lines to recreate the *first* line
(for the logical (D drive in this first Extended partition ):

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 2 0F 19535040 378748440 184935 1216 0 1 24791 254 63 OK
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

like so:
type: 0B, starting*: 1216 1 1, ending*: 8510 254 63, sectors before: 63, sectors: xxxxxxxxx,
where xxxxxxxxx = ((8511-1216)*255*63)-63 = 117194112
And with the same proviso for starting and ending CHS as before.

This will now be the D: logical but the bootrecord is likely toast.

If you can recreate your partition with same size on an other drive
and format it you could copy the bootrecord parameters from there
and input them here and see what gives. It could be that that may
give other apps like WinHEX enough to recognize directories and
such (if still there) and perhaps let you copy files and/or directories.


Another side note (to help better understand what we are doing above):

With microsoft there are usually 2 partitions: 1 primary, one extended.
You can see an extended partition as an description of "the rest".
Each extended can have another primary and another extended, which in
turn has .... and so on and so on. The last extended only has the last primary.
The primaries in the extendeds we call logicals.

The description of the extended partition serves as a link and the "sectors
before" are referenced against the start of the physical drive (sector 0).
"Sectors" are the remaining sectors on the physical drive.
Primaries are referenced logically against the start of the extended
partition (the EMBR), usually a track distance so usually 63 sectors.
"Sectors" here is the size of the logical drive.

 

[Joep]

That is what I prefer.



Possibly. I have no idea what that is about without further examination.

Lines copied from other post, for reference:

Disk: 1 Cylinders: 24792 Heads: 255 Sectors: 63 MB: 194474

Partitions according to partition bootrecords on first harddisk:
--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 - 0C 63 19534977 9538 0 1 1 1215 254 63 B OK
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

Partitions according to partition tables on first harddisk:
--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 1*0C 63 19534977 9538 0 1 1 1215 254 63 OK OK
0 2 0F 19535040 378748440 184935 1216 0 1 24791 254 63 OK

Punch the EMBR (extended partition MBR) button in the opening screen and
put most of the info of this line:

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

into line 2 of the new tables presented, like so:

type: 0F starting

You're now @ 19535040 and you're trying to 'define' the next EPBR right? (If
I am wrong it's due to the crappy instructions - also my other notes based
upon this assumption may be invalid as well). I think that needs to be 05
for type then. Never seen types 0F outside the MBR and I have seen a lot of
partition tables.

*: 8511 0 1, ending*: 24791 254 63, sectors before: xxxxxxxxx, sectors:
261554202,

where xxxxxxxxx = (8510+1)*(254+1)*63 = 136729215

The start LBA for the next EPBR should be relative to the start of the
extended or extendedX partition, at least that's how most tools do that and
because of that it is probably safer to do so as well. So 261554202 -
19535040 should be entered in the "Sectors Before' field.

One more side note. This sector (19535040) may be filled with garbage and
the lines in ptedit may contain garbage as well - It hardly ever is the case
that the sector was nicely wiped more often it was overwritten with
'garbage'. Ptedit will write the 55AA sig automatically if you click 'save
changes'..

(Cylinders and heads are 0-based)

* Starting and ending CHS will not be accepted for cylinders over 1023, you can fill
in the maximum CHS that is allowed for them: 1023 254 63, i.e. large drive

placeholders.

Ah! You got that right this time. BTW, to call that 'large drive
placeholders' is typically a PowerQuest thing. It's not so much that Ptedit
won't accept this but it is because for cylinder values only 10 bits are
available.

If all is well then ptedit should now let you go to the next EMBR for the

just defined

Click 'Save Changes' first.

(2nd) extended partition and show you the logical (E in it, as represented by:

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

and you should be able to open the bootrecord.

Next:
You can combine the info of these 2 lines to recreate the *first* line
(for the logical (D drive in this first Extended partition ):

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 2 0F 19535040 378748440 184935 1216 0 1 24791 254 63 OK
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

like so:
type: 0B, starting*: 1216 1 1, ending*: 8510 254 63, sectors before: 63, sectors: xxxxxxxxx,
where xxxxxxxxx = ((8511-1216)*255*63)-63 = 117194112
And with the same proviso for starting and ending CHS as before.

This will now be the D: logical but the bootrecord is likely toast.

If you can recreate your partition with same size on an other drive
and format it you could copy the bootrecord parameters from there
and input them here and see what gives. It could be that that may
give other apps like WinHEX enough to recognize directories and
such (if still there) and perhaps let you copy files and/or directories.


Another side note (to help better understand what we are doing above):

With microsoft there are usually 2 partitions: 1 primary, one extended.
You can see an extended partition as an description of "the rest".
Each extended can have another primary
and another extended, which in
turn has .... and so on and so on. The last extended only has the last primary.
The primaries in the extendeds we call logicals.

White ducks that are black, we call black ducks.

The description of the extended partition serves as a link and the "sectors
before" are referenced against the start of the physical drive (sector 0).
"Sectors" are the remaining sectors on the physical drive.
Primaries are referenced logically against the start of the extended
partition (the EMBR

(which themselves are relative to the start of the extended)),

 

[Annika K]

Here is my plan of attack...Please look it over and correct any
misunderstandings that I may have...

First of all, here is the First PTEdit screen before I make any changes....

Starting Ending Sectors
Type Boot Cyl Head Sector Cyl Head Sector Before Sectors
------------------------------------------------------------------
0C 80 0 1 1 191 254 63 63 19534977
0F 00 192 0 1 215 254 63 19535040 378748440

Then, after I click Goto EPBR the screen looks like this (before
modification)...

Starting Ending Sectors
Type Boot Cyl Head Sector Cyl Head Sector Before Sectors
------------------------------------------------------------------
FF 10 414 171 13 4 12 30 2565263962 1872013011
0D BC 737 20 54 586 155 24 715414953 2356450195
C1 9F 473 187 57 97 150 29 2960922799 89931260
51 33 158 72 17 302 240 1 886023102 2723627765

At this stage, I am supposed to change the second line so that the screen
looks like this...

Starting Ending Sectors
Type Boot Cyl Head Sector Cyl Head Sector Before Sectors
--------------------------------------------------------------------
FF 10 414 171 13 4 12 30 2565263962 1872013011
0F BC 1023 254 63 1023 254 63 19535040 261554202
C1 9F 473 187 57 97 150 29 2960922799 89931260
51 33 158 72 17 302 240 1 886023102 2723627765

....using CHS 1023,245,63 instead of the actual values because the cylinder
is over 1023 (is that correct? I'm unsure on this) and following Joep's
suggestions to use Type: 0F and "Sectors Before":19535040 (What about the
boot field? do I leave that BC?)

Then I click "Save Changes"
Then I click "Goto EPBR" for the 2nd line (the one i just modified)
Then I change the first line of the resulting screen to look like this

Starting Ending Sectors
Type Boot Cyl Head Sector Cyl Head Sector Before Sectors
--------------------------------------------------------------------
0B ?? 1023 254 63 1023 254 63 63 117194112

Again using CHS 1023,254,63 (What about the boot field here?)

Then I click Save changes and exit the program
Then i try booting the machine? I'm afraid

Thanks very much for your time in this matter.

 

[Folkert Rienstra]

You're now @ 19535040 and you're trying to 'define' the next EPBR right?

(If I am wrong it's due to the crappy instructions -

But not half as crappy as your posting style and crappy quoting.

also my other notes based upon this assumption may be invalid as well).

But not on that assumption.

I think that needs to be 05 for type then.

Apparently it doesn't hurt (or does it) as that is what I see as well.
As I understand it, the CHS values are ignored in extended-x (0F)
but not in extended (05). Since the values are placeholders anyway
0F would appear to be the right choice.

Never seen types 0F outside the MBR and I have seen a lot of partition tables.

Can't see why it would be a problem for an EPBR but not the MBR.
It appears however that some older OSes (DOS, Win95) can get confused by it.

The start LBA for the next EPBR should be relative to the start of the
extended or extendedX partition, at least that's how most tools do that
and because of that it is probably safer to do so as well.

If that is the way to do it then that is how it is done, period.
With values over 8GB the LBA values are the only values that
do remain to calculate addresses of the linked structures.

So 261554202 - 19535040 should be entered in the "Sectors Before' field.

Nope.
261554202 is the *size* of the last partition, not the physical start of that partition.

One more side note. This sector (19535040) may be filled with garbage and
the lines in ptedit may contain garbage as well - It hardly ever is the case
that the sector was nicely wiped more often it was overwritten with'garbage'.
Ptedit will write the 55AA sig automatically if you click 'save changes'..


Ah! You got that right this time.

BTW, to call that 'large drive placeholders' is typically a PowerQuest thing.
It's not so much that Ptedit won't accept this but it is because for cylinder
values only 10 bits are available.

You don't say. May be that is why Ptedit won't accept them then, don't you think?
And it is not that simple as Ptedit will also not allow 255 heads even though that
fits in the 8 bits reserved for it.
And some apps and/or OSes check these values to determine whether they will
see the drive as a smaller or *larger* than 8GB drive (whether that is correct
is an entirely different matter) and some even go in the modulo 32GB looparound.
So I think that "large drive placeholders" is a quite appropriate name for them.

Click 'Save Changes' first.

As if it will allow you there before saving (or discarding) any previously made changes.

[snip]

 

[Folkert Rienstra]

Here is my plan of attack...Please look it over and correct any misunderstandings
that I may have...

First of all, here is the First PTEdit screen before I make any changes....

Starting Ending Sectors
Type Boot Cyl Head Sector Cyl Head Sector Before Sectors

Hmm.

" > 0 1*0C 63 19534977 9538 0 1 1 1215 254 63 OK OK "
" > 0 2 0F 19535040 378748440 184935 1216 0 1 24791 254 63 OK "

Findpart seems to have a mind of it's own. Svend should look into that.

Then, after I click Goto EPBR the screen looks like this (before modification)...

Starting Ending Sectors
Type Boot Cyl Head Sector Cyl Head Sector Before Sectors
------------------------------------------------------------------
FF 10 414 171 13 4 12 30 2565263962 1872013011
0D BC 737 20 54 586 155 24 715414953 2356450195
C1 9F 473 187 57 97 150 29 2960922799 89931260
51 33 158 72 17 302 240 1 886023102 2723627765

At this stage, I am supposed to change the second line so that the screen
looks like this...

Starting Ending Sectors
Type Boot Cyl Head Sector Cyl Head Sector Before Sectors
--------------------------------------------------------------------
FF 10 414 171 13 4 12 0 2565263962 1872013011
0F BC 1023 254 63 1023 254 63 19535040 261554202
C1 9F 473 187 57 97 150 29 2960922799 89931260
51 33 158 72 17 302 240 1 886023102 2723627765

You can clear lines 3 and 4 if you want.

...using CHS 1023,245,63 instead of the actual values because the cylinder
is over 1023

(is that correct? I'm unsure on this)

Yes, that is correct. The "actual" values aren't really actual, they are con-
verted/recalculated (but invalid) CHS representations of the LBA values.
They don't actually exist and can't be actually used in bios or driver
routines because CHS doesn't exist beyond 8GB (1024*256*63*512).

and following Joep's suggestions to use Type: 0F and "Sectors Before":19535040

No, it should be my original value (136729215) but minus that 19535040 of the
first original primary which makes 117194175.

(What about the boot field? do I leave that BC?)

You may clear that.

Then I click "Save Changes"
Then I click "Goto EPBR" for the 2nd line (the one i just modified)

You don't have to, you just check that you can and then you return, or you can leave it.
And you don't have to position the cursor, Goto EPBR just goes to the next
EPBR in the chain, like Goto Parent will return you to the previous one.

Then I change the first line of the resulting screen to look like this

No, you return (Goto Parent) if you *did* go to the next EPBR to check if it's working.
As I said, the 3rd logical should be OK so you are not going to change anything there.
We will change the first line in the same screen where we changed that second line.

Starting Ending Sectors
Type Boot Cyl Head Sector Cyl Head Sector Before Sectors

You may clear it.
You may also clear any unused lines that have rubbish in them.

Then I click Save changes and exit the program
Then i try booting the machine? I'm afraid

Thanks very much for your time in this matter.

And it is taking more time than I expected it would. I must try and remember
how tricky it is and how easy you can make a mistake if you can't dry run it.
It's probably easier if you draw the screens on paper and check the
columns against each other than to try keeping it all in ones head.
Your constant removal of previous quotation doesn't make the process any
snappier either.

 

[Folkert Rienstra]

Time for another "oops".

That is what I prefer.



Possibly. I have no idea what that is about without further examination.

Lines copied from other post, for reference:

Disk: 1 Cylinders: 24792 Heads: 255 Sectors: 63 MB: 194474

Partitions according to partition bootrecords on first harddisk:
--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 - 0C 63 19534977 9538 0 1 1 1215 254 63 B OK
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

Partitions according to partition tables on first harddisk:
--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 1*0C 63 19534977 9538 0 1 1 1215 254 63 OK OK
0 2 0F 19535040 378748440 184935 1216 0 1 24791 254 63 OK

Punch the EPBR (Extended Partition mBR) button in the opening screen and
put most of the info of this line:

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

into line 2 of the new tables presented, like so:

type: 0F starting*: 8511 0 1, ending*: 24791 254 63, sectors before: xxxxxxxxx, sectors: 261554202,

where xxxxxxxxx = (8510+1)*(254+1)*63 = 136729215 (Cylinders and heads are 0-based)

So with the correction below in mind we have to correct for the space taken up by the primary partition:
136729215-19535040 = 117194175

* Starting and ending CHS will not be accepted for cylinders over 1023, you can fill
in the maximum CHS that is allowed for them: 1023 254 63, i.e. large drive placeholders.

If all is well then ptedit should now let you go to the next EPBR for the just defined
(2nd) extended partition and show you the logical (E in it, as represented by:

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

and you should be able to open the bootrecord.

Next:
You can combine the info of these 2 lines to recreate the *first* line
(for the logical (D drive in this first Extended partition ):

--PCyl N ID -----Rel -----Num ---MB --Start CHS- ---End CHS-- BS CHS
0 2 0F 19535040 378748440 184935 1216 0 1 24791 254 63 OK
8511 1 0B 63 261554202 127712 8511 1 1 24791 254 63 OK OK

like so:
type: 0B, starting*: 1216 1 1, ending*: 8510 254 63, sectors before: 63, sectors: xxxxxxxxx,
where xxxxxxxxx = ((8511-1216)*255*63)-63 = 117194112
And with the same proviso for starting and ending CHS as before.

This will now be the D: logical but the bootrecord is likely toast.

If you can recreate your partition with same size on an other drive
and format it you could copy the bootrecord parameters from there
and input them here and see what gives. It could be that that may
give other apps like WinHEX enough to recognize directories and
such (if still there) and perhaps let you copy files and/or directories.


Another side note (to help better understand what we are doing above):

With microsoft there are usually 2 partitions: 1 primary, one extended.
You can see an extended partition as a description of "the rest".
Each extended can have another primary and another extended, which in
turn has .... and so on and so on. The last extended only has the last primary.
The primaries in the extendeds we call logicals.

The description of the extended partition serves as a link and the "sectors
before" are referenced against the start of the physical drive (sector 0).

Oops, that's wrong, It's relative to a single point, but it is not sector 0.
It's relative to the start of the original extended partition, the disk address
of the first EPBR.

 

[Joep]

But not half as crappy as your posting style and crappy quoting.

Crappy, crappy, crappy!

But not on that assumption.

That assumption plus the crappy instructions.

Apparently it doesn't hurt (or does it).

Aarghhhh, it hurts!

as that is what I see as well.
As I understand it, the CHS values are ignored in extended-x (0F)
but not in extended (05). Since the values are placeholders anyway
0F would appear to be the right choice.

No it wouldn't.

tables.

Can't see why it would be a problem for an EPBR but not the MBR.
It appears however that some older OSes (DOS, Win95) can get confused by it.

If that is the way to do it then that is how it is done, period.
Period.

With values over 8GB the LBA values are the only values that
do remain to calculate addresses of the linked structures.


Nope.
261554202 is the *size* of the last partition, not the physical start of

that partition.

Well, now you see how crappy your instructions are.

You don't say.
Yup.

May be that is why Ptedit won't accept them then, don't you think?

Yup. That's what I said.

And it is not that simple as Ptedit will also not allow 255 heads even though that
fits in the 8 bits reserved for it.

That's to work around a DOS and every OS that's build on it limitation.

And some apps and/or OSes check these values to determine whether they will
see the drive as a smaller or *larger* than 8GB drive (whether that is correct
is an entirely different matter) and some even go in the modulo 32GB looparound.
So I think that "large drive placeholders" is a quite appropriate name for

them.

I never argued that it isn't.

As if it will allow you there before saving (or discarding) any previously

made changes.

That's why you need to click it.

 

[Joep]

And it is taking more time than I expected it would. I must try and remember
how tricky it is and how easy you can make a mistake if you can't dry run it.
It's probably easier if you draw the screens on paper and check the
columns against each other than to try keeping it all in ones head.
Your constant removal of previous quotation doesn't make the process any
snappier either.

Not trying to be a smart-ass (it's hard) I'd like to make one suggestion:

Re-type your instructions from scratch so there's one nicely formatted email
with only correct values.

 

[Tanmoy]

Hi... If you had installed Windows 98 on this HD you should be able to
see 2 files SUHODLOG.DAT and SUHODLOG.BAK in rott of C:\ these are
hidden system files. Send me these two files and I will send you your
original Partition table along with a restoration utility developed by
me.

Best regards
Tanmoy Banerjee

 

[Joep]

Hi... If you had installed Windows 98 on this HD you should be able to
see 2 files SUHODLOG.DAT and SUHODLOG.BAK in rott of C:\

Ah! in the rott! Splendid! And isn't that SUHDLOG.DAT?

these are
hidden system files. Send me these two files and I will send you your
original Partition table

Partition locations are already known. Read the rest of the thread.

 

[Folkert Rienstra]

Not trying to be a smart-ass (it's hard)

And failing.

I'd like to make one suggestion:

Re-type your instructions from scratch so there's one nicely formatted email
with only correct values.

Well, what holds you back to do that yourself and take the credit?