bad sectors on my ntfs partition, what do i do?

[Haneef Bashir]

i have a three partitioned harddrive. i have windows on the first 10gb
partition. after having some difficulties with windows and the
reinstallation of windows, and now not being able to boot back into
hardy, i grabbed my copy of puppy linux and used the live-cd (which i
am typing on now). i then used gparted to check and repair the windows
partition and this is what i got:
<i>ntfsresize v2.0.0 (libntfs 10:0:0)<BR />Device name :
/dev/sda1<BR />NTFS volume version: 3.1<BR />Cluster size : 4096
bytes<BR />Current volume size: 11202798080 bytes (11203 MB)<BR
/>Current device size: 11202799104 bytes (11203 MB)<BR />Checking for
bad sectors ...<BR />Bad cluster: 0xbd0bc - 0xbd0bc (1)<BR />Bad
cluster: 0xbd1f7 - 0xbd1f7 (1)<BR />Bad cluster: 0xbd306 -
0xbd306 (1)<BR />Bad cluster: 0xc0029 - 0xc0029 (1)<BR />ERROR:
This software has detected that the disk has at least 4 bad
sectors.<BR />****************************************************************************<BR
/>* WARNING: The disk has bad sector. This means physical damage on
the disk *<BR />* surface caused by deterioration, manufacturing
faults or other reason. *<BR />* The reliability of the disk may
stay stable or degrade fast. We suggest *<BR />* making a full backup
urgently by running &apos;ntfsclone --rescue ...&apos; then *<BR />*
run &apos;chkdsk /f /r&apos; on Windows and rebooot it TWICE! Then you
can resize *<BR />* NTFS safely by additionally using the
--bad-sectors option of ntfsresize.*<BR
/>****************************************************************************<BR
/></i>

i just want to know is this haddrive still any good? can it be fixed,
can i install windows on another partition with out any problem?

--
ubuntu-users mailing list
ubuntu...@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-users

[Karl F. Larsen]

Read what your computer said and back up what you want and then
throw away that hard drive. It's dead.

karl

[Haneef Bashir]

>    Read what your computer said and back up what you want and then
> throw away that hard drive. It's dead.
>

ok is there a way to save my ubuntu install?

[Karl F. Larsen]

Haneef Bashir wrote:
>> Read what your computer said and back up what you want and then
>> throw away that hard drive. It's dead.
>>
>>
> ok is there a way to save my ubuntu install?
>
>

What do you mean by your ubuntu install? Trust nothing on that hard
drive! Save files you want to save and hope they are complete.

73 Karl

[Brian McKee]

On Sun, Aug 23, 2009 at 12:17 PM, Haneef Bashir<darkf...@gmail.com> wrote:
> i just want to know is this haddrive still any good? can it be fixed,
> can i install windows on another partition with out any problem?

Harddrives can reassign bad sectors automatically - when they start
showing up without the harddrive hiding them from you before you got
to them it's often not a good sign.

Step one is make darned sure you have a backup of everything on that drive, then

You could run Spinrite on it - a commercial product that attempts to
recover the data from the bad partitions and rewrite them elsewhere.
A pass with it will also give you a better idea of just how many
issues there are on the drive. Run it a second time and if new
problems keep popping up it's time to trash the drive. If the
problems stay fixed and the SMART data looks ok, than I'd put it back
in service.

If you'd prefer free or open solutions, you could repartition it to a
single partition, then run mke2fs -cc on it - that will do a
read-write test for bad blocks as it formats. Once it completes I'd
write /dev/random to the whole drive with dd as a last test watching
the output for errors, again if it looks ok and SMART tests pass, I'd
put it back in service.l

HTH
Brian

--
All you need to know about Ubuntu 9.04 Jaunty -> gconftool -s --type
bool /apps/update-notifier/auto_launch false

[Haneef Bashir]

On Mon, Aug 24, 2009 at 3:43 AM, Brian McKee<brian...@gmail.com> wrote:
> On Sun, Aug 23, 2009 at 12:17 PM, Haneef Bashir<darkf...@gmail.com> wrote:
>> i just want to know is this haddrive still any good? can it be fixed,
>> can i install windows on another partition with out any problem?
>
> Harddrives can reassign bad sectors automatically - when they start
> showing up without the harddrive hiding them from you before you got
> to them it's often not a good sign.
>
> Step one is make darned sure you have a backup of everything on that drive, then
>
> You could run Spinrite on it - a commercial product that attempts to
> recover the data from the bad partitions and rewrite them elsewhere.
> A pass with it will also give you a better idea of just how many
> issues there are on the drive.  Run it a second time and if new
> problems keep popping up it's time to trash the drive.  If the
> problems stay fixed and the SMART data looks ok, than I'd put it back
> in service.
>
> If you'd prefer free or open solutions, you could repartition it to a
> single partition, then run mke2fs -cc on it - that will do a
> read-write test for bad blocks as it formats.  Once it completes I'd
> write /dev/random to the whole drive with dd as a last test watching
> the output for errors, again if it looks ok and SMART tests pass, I'd
> put it back in service.l
>

thanks brian ill try that out.

[jdow]

The disk is dead, Jim er Haneef. You need to rescue it.

You have linux. That's a really good start.

Use it to make a duplicate of the whole disk containing the bad
sectors, using dd.

Copy using:
"dd bs=512 CONV=sync,fdatasync,noerror if=/dev/sdb of=/dev/sdc"
Modify "sdb" and "sdc" as appropriate. Note that you MAY have to
adjust the flags for dd to make it happy and make the output copy
size the same as the number of disk blocks reported via something
like "fdisk -l /dev/sdb".

(This will be brutally slow. But every block that can be read will be
read and placed on the output disk properly. Double check that dd
declares the correct number of blocks was copied.)

Then mount the disk with XP and let it churn until the disk is fixed.
Run a check disk on it to be sure.

I used that to repair a disk that the real NTFS could not fix because
there was a directory block that could not be repaired or written fairly
deep in the system files. Luckily when the smoke cleared the duplicate
was able to be fixed and all I lost were some of the system restore
files from "way back" that I didn't nead anyway.

You might be able to use ntfsclone as indicated. But I'd prefer to use
it on a duplicate, though. Of course, that would involve three disks.
So use your own imagination for how you want to arrange it all. My
personal experience is with this "dd" technique that worked like a
serious champ for me with a disk that would not even boot it's XP
image because of the broken directory block. It's image is now booting
XP on my laptop when I need it. (I used the extra space on the new disk
for another storage partition for XP. I have a separate Linux disk I use.)

{^_^}

[Rashkae]

Brian McKee wrote:

>
> If you'd prefer free or open solutions, you could repartition it to a
> single partition, then run mke2fs -cc on it - that will do a
> read-write test for bad blocks as it formats. Once it completes I'd
> write /dev/random to the whole drive with dd as a last test watching
> the output for errors, again if it looks ok and SMART tests pass, I'd
> put it back in service.l
>

I agree with most everything said, but for this step, the right tool for
the right job is badblocks.

Once your certain all data has been copied from the potentially failing
drive, run badblocks with either the -n or -w switch. (-n is
theoretically non-destructive, if the drive isn't in non-functioning
state and the process completes uninterrupted. -w is complete
destruction of the drive, and will be equivalent of a low level format
as well as complete surface test.)

[Haneef Bashir]

On Mon, Aug 24, 2009 at 9:27 PM, Rashkae<ubu...@tigershaunt.com> wrote:
> Brian McKee wrote:
>
>>
>> If you'd prefer free or open solutions, you could repartition it to a
>> single partition, then run mke2fs -cc on it - that will do a
>> read-write test for bad blocks as it formats.  Once it completes I'd
>> write /dev/random to the whole drive with dd as a last test watching
>> the output for errors, again if it looks ok and SMART tests pass, I'd
>> put it back in service.l
>>
>
> I agree with most everything said, but for this step, the right tool for
> the right job is badblocks.
>
> Once your certain all data has been copied from the potentially failing
> drive, run badblocks with either the -n or -w switch.  (-n is
> theoretically non-destructive, if the drive isn't in non-functioning
> state and the process completes uninterrupted.  -w is complete
> destruction of the drive, and will be equivalent of a low level format
> as well as complete surface test.)

ok ill try that. thanks all for the help.

[Brian McKee]

2009/8/24 Rashkae <ubu...@tigershaunt.com>:

> Brian McKee wrote:
>> If you'd prefer free or open solutions, you could repartition it to a
>> single partition, then run mke2fs -cc on it - that will do a
>> read-write test for bad blocks as it formats.
>

> I agree with most everything said, but for this step, the right tool for
> the right job is badblocks.

Then we really agree - mke2fs -cc calls badblocks :-)

Brian

--
All you need to know about Ubuntu 9.04 Jaunty -> gconftool -s --type
bool /apps/update-notifier/auto_launch false

--