Bad Sector Retesting tool for NTFS?

  • Thread starter Thread starter Sheridan Hutchinson
  • Start date Start date
S

Sheridan Hutchinson

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,
I had a previous hard disk that had 4k worth of bad sectors. When I
got my nice shiny new IBM replacement I cloned the entire thing
across and resized the partition, but of course, it took with it the
MFT and the "record" of the bad sectors.

How do I get XP to retest the sectors that have been marked out as
bad but aren't? Is there any 3rd party tool anyone knows of that
could do this?

- --
Regards,
Sheridan Hutchinson
(e-mail address removed) or ICQ# 332-123-498


-----BEGIN PGP SIGNATURE-----
Version: PGP 8.0.3

iQA/AwUBQIS1Mfw9TKasPKYwEQIRJACgr5L8jJlsh6++ABzKDC92cZCT+LMAnR5O
8KC3sq/zy/FFheNzwyZ49dT5
=NjnO
-----END PGP SIGNATURE-----
 
PartitionMagic can do a bad sector re-test.
Click to expand...


It seems to me that the bad sector table is hidden below the file
system. If a block becomes marginal (ECC and retry get good data but
exceed some threashold) the data is copied to a spare block, and the
marginal sector is added to a list you can't see without running
special diagnostic tools. Is there something I'm missing ?

If a sector suddenly goes completely bad then you will be able to
copy every file except the one with the bad block. XCOPY /C should
bypass any file that can't be read, except that I'be seen a disk
retry endlessly and effectivly hang even with /C.

If a disk is seriously degrading it's possible to fill up the spare
sector table.
 
Al Dykes said:
It seems to me that the bad sector table is hidden below the file
system. If a block becomes marginal (ECC and retry get good data but
exceed some threashold) the data is copied to a spare block, and the
marginal sector is added to a list you can't see without running
special diagnostic tools. Is there something I'm missing ?
Click to expand...

Yes, you are indeed missing something. Of course many drives come with an
internal defect management which can take a sector out of service and
replace it with a 'new' one (from a spare pool).

But that's not what OP is talking about: NTFS also keeps a bad block table,
and, after copying a disk to a *new disk the same blocks are flagged as
'bad'. So you may want to tell NTFS that *on this new disk* the areas maked
bad really aren't. PartitionMagic can do this.
 
How do I get XP to retest the sectors that have been marked out as
bad but aren't?
Click to expand...

Open "My Computer" and right click on the drive in question. Select
Properties, then Tools, then Error-checking. There is an option to
check for and recover bad sectors. You will need to reboot if this is
the system drive.

You can also boot from the XP install disks, select repair and them
select command line. Run chkdsk /? for the options avaliable. As I
recall, scan for and recover bad sectors is an option.

- -
Gary L.
Reply to the newsgroup only
 
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Gary L. said:
Open "My Computer" and right click on the drive in question. Select
Properties, then Tools, then Error-checking. There is an option to
check for and recover bad sectors. You will need to reboot if this
is the system drive.

You can also boot from the XP install disks, select repair and them
select command line. Run chkdsk /? for the options avaliable. As I
recall, scan for and recover bad sectors is an option.
Click to expand...

Hi, thanks for the advice, I had already previously checked out
chkdsk and the inbuilt windows diagnostic tool, but neither provide
an option of retesting the bad sector map; they only look for new
ones.

- --
Regards,
Sheridan Hutchinson
(e-mail address removed) or ICQ# 332-123-498


-----BEGIN PGP SIGNATURE-----
Version: PGP 8.0.3

iQA/AwUBQIwb2/w9TKasPKYwEQIDfgCfd9zH0XhpfepzSm3J968KQ+HrifUAn0xO
8MHi1VVTtmmUxPI8Co6ET968
=vtT0
-----END PGP SIGNATURE-----
 
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Joep said:
PartitionMagic can do a bad sector re-test.
Click to expand...

Hi, thanks for your advice!

I did check this option out, unfortunately the test can only be
executed on a FAT partition :(
- --

Regards,
Sheridan Hutchinson
(e-mail address removed) or ICQ# 332-123-498


-----BEGIN PGP SIGNATURE-----
Version: PGP 8.0.3

iQA/AwUBQIwc2fw9TKasPKYwEQJ+9ACdHW5O22p4b0yfWsiog4E8mKBraXcAoLv0
bDQZqqVh9wXZw5AHbVbrfv1Q
=us9v
-----END PGP SIGNATURE-----
 
You got me there ;-)
Click to expand...

If you don't have any data on the disk the type of file system
shouldn't matter.

Can you blow the data away and use a low-level utility to update the
bad block table. Once you have mapped out the marginal secters are
remapped to good ones and a blockcopy utility like Ghost should work
becase the target disk looks perfect at that level.

I've booted a knoppix disk and run badblocks with several passed
to sniff out bad sectors.
 
Al Dykes said:
If you don't have any data on the disk the type of file system
shouldn't matter.

Can you blow the data away and use a low-level utility to update the
bad block table.
Click to expand...

And still he has no idea of what bad block table is referred to.
There are no bad sectors on the target! They were on his previous harddrive.
Once you have mapped out the marginal secters are remapped to good
ones and a blockcopy utility like Ghost should work becase the target
disk looks perfect at that level.
Click to expand...

The target drive is perfect already (hopefully).
Still will copy the not used socalled bad sectors in the source to perfectly
good sectors in the target where the filesystem will say that they are not used.
He wants them back in use (all 4k of them).
I've booted a knoppix disk and run badblocks with several passed
to sniff out bad sectors.
Click to expand...

The 'socalled' bad sectors are in the image, get it?
 
If you don't have any data on the disk the type of file system
shouldn't matter.
Click to expand...

Again, you don't get it, READ first then waste bandwith! There IS data on
the disk why would one CLONE it if there wasn't? Since the $badclus meta
file is part of the NTFS file system, the file system DOES matter in this
case.

I KNOW you can have 'bad' sectors taken out of service and have them
replaced with spares (my own tool HD Workbench does this). However OP stated
NTFS recorded bad clusters in the $badclus meta file and it stays this way
EVEN when you take care of bad sectors on a lower level.

Also, OP already CLONED the disk, so bad sectors that existed on the source
do not exist on the DESTINATION, however they're still flagged as bad in the
$badclus meta file. If $badclus ALREADY contains references to bad clusters
then you are TO LATE to take care of them on a lower level!!!

BTW: Ghost is not a block copy utility.
 
Again, you don't get it, READ first then waste bandwith! There IS data on
the disk why would one CLONE it if there wasn't? Since the $badclus meta
file is part of the NTFS file system, the file system DOES matter in this
case.

I KNOW you can have 'bad' sectors taken out of service and have them
replaced with spares (my own tool HD Workbench does this). However OP stated
NTFS recorded bad clusters in the $badclus meta file and it stays this way
EVEN when you take care of bad sectors on a lower level.

Also, OP already CLONED the disk, so bad sectors that existed on the source
do not exist on the DESTINATION, however they're still flagged as bad in the
$badclus meta file. If $badclus ALREADY contains references to bad clusters
then you are TO LATE to take care of them on a lower level!!!

BTW: Ghost is not a block copy utility.
Click to expand...

I said "if"....

If you are copying a filesystem with an NTFS-aware tool why would
it copy the metafile ? It's clearly never the right thing to do.
 
Al Dykes said:
I said "if"....
Click to expand...

Okay, then it's me, it confused me because you already made that point.
If you are copying a filesystem with an NTFS-aware tool why would
it copy the metafile ? It's clearly never the right thing to do.
Click to expand...

This is true. I can not think of circumstances where it would be usefull to
copy the $badclus file. However, if not copied the $bitmap needs to be
updated as well as the bad clusters are flagged to be 'in use'. It'd be
probably a good idea if Symantec added this.
 
Okay, then it's me, it confused me because you already made that point.


This is true. I can not think of circumstances where it would be usefull to
copy the $badclus file. However, if not copied the $bitmap needs to be
updated as well as the bad clusters are flagged to be 'in use'. It'd be
probably a good idea if Symantec added this.
Click to expand...

Now we're getting somewhere. I'd argue that you don't copy
the bitmap file, either, if the tool is NTFS-aware.


I'm going to spend time browsing

http://www.ntfs.com/
 
Joep said:
This is true. I can not think of circumstances where it would be usefull to
copy the $badclus file. However, if not copied the $bitmap needs to be
updated as well as the bad clusters are flagged to be 'in use'. It'd be
probably a good idea if Symantec added this.
Click to expand...
The $bitmap file cannot be copied as the source and dest are usually different
sizes. I doubt if Ghost copies the $badclus file. What product did the OP use?
 
Think it was PM. Where PM is a sector oriented tool orignally, Ghost isn't.
So they will probably handle this in a different manner.

I can imagine you'd still copy the $bitmap even when resizing a partition on
the fly, all you'd have to do is modify the $bitmap to reflect the new size.
I'm not sure but I can imagine the $bitmap is a continguous file, so you may
need to reallocate it as well.
 
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Eric Gisin said:
The $bitmap file cannot be copied as the source and dest are
usually different sizes. I doubt if Ghost copies the $badclus file.
What product did the OP use?
Click to expand...

I've still had no joy with this one yet :(

I used Partition Magic, which uses a sector based copy which is why
the bad cluster map was copied as well.

You see, I thought there would be an easy fix for this, just some low
level tool that could sort it out, given how many people this
probably effects. I will find a way!

- --
Regards,
Sheridan Hutchinson
(e-mail address removed) or ICQ# 332-123-498


-----BEGIN PGP SIGNATURE-----
Version: PGP 8.0.3

iQA/AwUBQJAWT/w9TKasPKYwEQL18gCfcm4Qhp4q5UiJ/ScN1VLpNlz8lasAoLbH
km7QbxyLbp6OPYe7/1mD/xbG
=n0CI
-----END PGP SIGNATURE-----
 
Sheridan Hutchinson said:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,
I had a previous hard disk that had 4k worth of bad sectors. When I
got my nice shiny new IBM replacement I cloned the entire thing
across and resized the partition, but of course, it took with it the
MFT and the "record" of the bad sectors.

How do I get XP to retest the sectors that have been marked out as
bad but aren't? Is there any 3rd party tool anyone knows of that
could do this?
Click to expand...

The only way to reset NTFS bad block list is explained at
http://www.bodden.de/misc/ntfsrecovery/

Partition Magic, Windows' chkdsk, etc can't do it, as you also already
experienced.
 
Back
Top