Chkdsk problem...

[Louis]

Hello,

While trying to make an image of an NTFS partition, the imaging
softwares (I tried with both "True Image" and "Ghost") reported that
errors on partition prevented the image to be completed.

For example, Acronis stated that it "Failed to read from the sector
234,439,534 of the hard disk 1", which seems high to me since there's
only 29,304,934 clusters on the disk.

Back into XP, I performed a chkdsk (without /F) from a cmd window, and
it reported it found "minor" errors.

I reentered the same command, this time with the /F switch, and a scan
was planned on next boot.

In this scan. chkdsk performed steps 1 to 3 and USN log without
reporting any error, but failed right after with a message "Unable to
write the second NTFS boot sector" (translation of the french message)
before going on with normal windows boot.

I then permormed "chkdsk c: /P /R" from the recovery console.

This time, chkdsk progress raised up to 75% then, felt to 50% before
raising again (slowly) to 75% - without any additional message. This
seems to happens a couple of times.

At some point, it stopped with the message "The volume seems to contain
one or many unrecoverable problems".

Under normal windows operation, nothing seems to indicate a problem
with the file system. All programs seem to run normally. Even a
complete virus scan (accessing the over 300,000 files on the disk)
doesn't show any sign of failing disk. I know it doesn't mean all my
files are in safe state, but at least they are not causing system
crashes or error messages.

While I already have a backup of all my data files, I'd like to see if
there's a way to repair that partition to a point where I'll be able to
image it.

BTW, the disk is a WD-1200JB, S.M.A.R.T doesn't report any error.

Any suggestions ?
Thanks

PS: I retried the console's chkdsk again this morning, and noted that
the 50% ... -> 75% -> 50% ...75% seems to occur more often than the
first time I made it. Does this means it is getting any nearer to
complete recovery ?

 

[Rod Speed]

While trying to make an image of an NTFS partition, the imaging
softwares (I tried with both "True Image" and "Ghost") reported that errors on partition prevented
the image to be completed.

For example, Acronis stated that it "Failed to read from the sector 234,439,534 of the hard disk
1", which seems high to me since there's only 29,304,934 clusters on the disk.

Sectors arent clusters. How may sectors per cluster do you have ?
That sector number would be appropriate for a drive bigger than 120G.

Back into XP, I performed a chkdsk (without /F) from a cmd window, and it reported it found
"minor" errors.

That wont say anything useful about a bad sector at 234,439,534

I reentered the same command, this time with the /F switch, and a scan was planned on next boot.

In this scan. chkdsk performed steps 1 to 3 and USN log without
reporting any error, but failed right after with a message "Unable to
write the second NTFS boot sector" (translation of the french message)
before going on with normal windows boot.

I then permormed "chkdsk c: /P /R" from the recovery console.

This time, chkdsk progress raised up to 75% then, felt to 50% before
raising again (slowly) to 75% - without any additional message. This
seems to happens a couple of times.

At some point, it stopped with the message "The volume seems to contain one or many unrecoverable
problems".

Under normal windows operation, nothing seems to indicate a problem with the file system. All
programs seem to run normally.

All that means is that the bad sectors arent being used in normal ops.

Even a complete virus scan (accessing the over 300,000 files on the disk) doesn't show any sign of
failing disk. I know it doesn't mean all my files are in safe state, but at least they are not
causing system crashes or error messages.
Yet.

While I already have a backup of all my data files, I'd like to see if there's a way to repair
that partition to a point where I'll be able to image it.

BTW, the disk is a WD-1200JB, S.M.A.R.T doesn't report any error.

Any suggestions ?

Post an Everest SMART report.
http://www.majorgeeks.com/download.php?det=4181

PS: I retried the console's chkdsk again this morning, and noted that the 50% ... -> 75% -> 50%
...75% seems to occur more often than the first time I made it. Does this means it is getting
any nearer to complete recovery ?

Nope, its more likely that the bad sectors vary in readability.

 

[Louis]

Rod Speed vient de nous annoncer :

Sectors arent clusters. How may sectors per cluster do you have ?

8

I calculated 8 * 29,304,934 = 234,439,472, but forgot to take into
account that the partition doesn't start at sector 0.

Actually, the last partition's sector is 234,439,535.

Post an Everest SMART report.

----[ SMART ]-----------------------------------------------------

[ WDC WD1200JB-00CRA1 (WD-WMA8C4206690) ]

01 Raw Read Error Rate 51 200 200 0 OK
03 Spin Up Time 21 97 93 5916 OK
04 Start/Stop Count 40 98 98 2697 OK
05 Reallocated Sector Count 140 200 200 0 OK
07 Seek Error Rate 51 200 200 0 OK
09 Power-On Time Count 0 91 91 6679 OK
0A Spin Retry Count 51 100 100 0 OK
0B Calibration Retry Count 51 100 100 0 OK
0C Power Cycle Count 0 98 98 2652 OK
C4 Reallocation Event Count 0 200 200 0 OK
C5 Current Pending Sector Count 0 200 200 0 OK
C6 Off-Line Uncorrectable Sector Count 0 200 200 0 OK
C7 Ultra ATA CRC Error Rate 0 200 253 0 OK
C8 Write Error Rate 51 200 200 0 OK

As you surely know, on the right of the "OK" status is text saying
"Normal value" or "Will always work".

 

[Louis]

Just downloaded Data Lifeguard Diagnostics 5.04f from WD site,
performed both quick and extended tests and they both ended with
"status code 0000: no errors found".

 

[Rod Speed]

Rod Speed wrote

I calculated 8 * 29,304,934 = 234,439,472, but forgot to take into account that the partition
doesn't start at sector 0.

What really matters is the datasheet value of the number of sectors.
http://wdc.custhelp.com/cgi-bin/wdc...WFyY2hfdGV4dD1XRDEyMDBKQg**&p_li=&p_topview=1
says 234,441,648 sectors.

Actually, the last partition's sector is 234,439,535.

Looks like it may have got partitioned with the wrong
geometry and that is what confused the imagers.

----[ SMART ]-----------------------------------------------------

[ WDC WD1200JB-00CRA1 (WD-WMA8C4206690) ]

01 Raw Read Error Rate 51 200 200 0 OK
03 Spin Up Time 21 97 93 5916 OK
04 Start/Stop Count 40 98 98 2697 OK
05 Reallocated Sector Count 140 200 200 0 OK
07 Seek Error Rate 51 200 200 0 OK
09 Power-On Time Count 0 91 91 6679 OK
0A Spin Retry Count 51 100 100 0 OK
0B Calibration Retry Count 51 100 100 0 OK
0C Power Cycle Count 0 98 98 2652 OK
C4 Reallocation Event Count 0 200 200 0 OK
C5 Current Pending Sector Count 0 200 200 0 OK
C6 Off-Line Uncorrectable Sector Count 0 200 200 0 OK
C7 Ultra ATA CRC Error Rate 0 200 253 0 OK
C8 Write Error Rate 51 200 200 0 OK

That doesnt show any problems.

As you surely know, on the right of the "OK" status is text saying "Normal value" or "Will always
work".

Its the data value before the OK that matters a lot more.

Just downloaded Data Lifeguard Diagnostics 5.04f from WD site, performed both quick and extended
tests and they both ended with "status code 0000: no errors found".

Yeah, the drive is fine.

 

[Louis]

Rod Speed avait énoncé :

Looks like it may have got partitioned with the wrong
geometry and that is what confused the imagers.

According to the BIOS (Award) the drive (set to AUTO) has :

Cylinder 57460
Head 16
Precomp 0
Landing Zone 57459
Sector 255

Access Mode Auto

That doesnt show any problems.


Yeah, the drive is fine.

Ok. So I have a fine drive, which doesn't show any problem in normal
use other than windows unable to complete a simple chkdsk on it and
utilities such as Ghost, Acronis True Image, Paragon Partition Manager
and even Partition Magic (yeerk) unable to handle correctly..


Any other suggestion on what's happening here and how to increase
chances to successfully image that partition ?

 

[Rod Speed]

Rod Speed wrote

According to the BIOS (Award) the drive (set to AUTO) has :

Cylinder 57460
Head 16
Precomp 0
Landing Zone 57459
Sector 255

Access Mode Auto

Its either auto, or those value have been used. You cant have it both ways.

Ok. So I have a fine drive, which doesn't show any problem in normal
use other than windows unable to complete a simple chkdsk on it and
utilities such as Ghost, Acronis True Image, Paragon Partition Manager
and even Partition Magic (yeerk) unable to handle correctly..

If they are all having a problem with the drive, the simplest approach
is to copy everything off, set the drive type to auto, wipe the drive and
recreate the partition and copy everything back. Not that easy to do
if you cant image the drive tho.

Any other suggestion on what's happening here

It may have been formatted with different geometry than what is being used now.

and how to increase chances to successfully image that partition ?

See above. You could use something like xxclone which doesnt use the geometry data,
but you would need another drive of the same size as the used space to do that with.

 

[Louis]

Any other suggestion on what's happening here

It may have been formatted with different geometry than what is being used
now.

Okay, but why can't chkdsk handle this ?

 

[Louis]

Rod Speed a pensé très fort :

Its either auto, or those value have been used. You cant have it both ways.

I guess you understand that these values are those set, detected (and
reported) by "auto" mode ?

 

[Rod Speed]

Okay, but why can't chkdsk handle this ?

It was never designed to do that. It just fixes problems that develop over time
with partitions that have been formatted with the geometry they currently have.

 

[Rod Speed]

Rod Speed wrote

I guess you understand that these values are those set, detected (and
reported) by "auto" mode ?

Surely you mean the auto detect entry. Thats not the same as the AUTO drive type.

 

[Louis]

Rod Speed a formulé la demande :

It was never designed to do that. It just fixes problems that develop over
time
with partitions that have been formatted with the geometry they currently
have.

Okay.. So there's nothing I can do other than perform a backup
(NTBackup), wipe that drive and start over ?

Does NTBackup take care of all the system files, registry, etc ?

 

[Rod Speed]

Rod Speed wrote

Okay.. So there's nothing I can do other than perform a backup (NTBackup), wipe that drive and
start over ?

Its likely possible to fix it some other way, but I havent tried.

You could try Acronis Disk Director Suite and see if it will clean the mess up.
But you should have a full backup before trying anything, so you might as well
just delete the partition and replace it with the correct drive type after wiping anyway.

Does NTBackup take care of all the system files, registry, etc ?

Yes, but its a bit clumsy compared with say xxclone.

 

[Horst Franke]

I then permormed "chkdsk c: /P /R" from the recovery console.
This time, chkdsk progress raised up to 75% then, felt to 50% before
raising again (slowly) to 75% - without any additional message. This
seems to happens a couple of times.

At some point, it stopped with the message "The volume seems to
contain one or many unrecoverable problems".

OK, Louis, I would treat this as a _real problem_!
Those things with felt back to 50 % are known by me for my HDD too!
This does not only look like logical problems but also like faulty sectors!

I also have such a device under Win98SE. On Start Up it sometimes take
TWO hours to recover from the faulty sectors (complete the Win98SE Boot)!

Under normal windows operation, nothing seems to indicate a problem
with the file system. All programs seem to run normally. Even a
complete virus scan (accessing the over 300,000 files on the disk)
doesn't show any sign of failing disk. I know it doesn't mean all my
files are in safe state, but at least they are not causing system
crashes or error messages.

Yes that's same for me. NO, a Virus Scan has NOTHING to do with the
HDD consistency! Logical or sector failures are BESIDES any virus!

While I already have a backup of all my data files, I'd like to see if
there's a way to repair that partition to a point where I'll be able
to image it.

No problem, just delete the partition and recreate it.

But be aware, before that operation there should also be something
like a CHKDSK verification!
Even more, many partition creators implement a previous CHKDSK
in any form! So to be on a save way, DO it from Win before!
If there is a faulty area, then try to skip it by assigning it as UNUSED!

BTW, the disk is a WD-1200JB, S.M.A.R.T doesn't report any error.

Ok, I don't know how effective the SMART report is.
Have too less experience (not enough years to verify) about SMART.

Any suggestions ?

You have a problem with Your disk! You need to fix this before any
further actions!
Every partition change requires this check! Even any SW to create
a partition image!

How should a partition be saved when it is logically corrupted?
That means CHKDSK MUST run clean, before any of those actions!

retried the console's chkdsk again this morning, and noted that
the 50% ... -> 75% -> 50% ...75% seems to occur more often than the
first time I made it. Does this means it is getting any nearer to
complete recovery ?

Sure! It says there ARE problems with Your disk!
If You have a complete backup of that partition, then delete and recreate
it.

For MY specific case, I would have to mark a certain area as faulty, mark
it as UNUSABLE and move the new system partition AFTER that area.

Otherwise any Win98SE startup would need hours to recover that errors
and after reBoot it's again the same situation!
For me it's the boot-partition, which needs to be moved!

At the moment I tend to operate this HDD on a second PC in order
to perform the cleanup from there (HDD physically moved).
Horst

 

[Horst Franke]

8
I calculated 8 * 29,304,934 = 234,439,472, but forgot to take into
account that the partition doesn't start at sector 0.
Actually, the last partition's sector is 234,439,535.

Louis, so Your calculation was faulty!
Never mind, there is some overhead to be taken into account!
But last sector *535 is higher than *534!
But as this is very very near to the limit, it could also be a problem
of Acronis!

Have You already reported this to Acronis?
What Version of what special product have You used?
Horst

 

[Horst Franke]

And there is some more overhead (234.441.648 - 234.439.535) which
is about 2113 sectors (may be related to NTFS + some more)?

What really matters is the datasheet value of the number of sectors.
http://wdc.custhelp.com/cgi-bin/wdc...WFyY2hfdGV4dD1XRDEyMDBKQg**&p_li=&p_topview=1
says 234,441,648 sectors.

NO, the datasheets like above tells of
User Sectors Per Drive 234,441,648
which is above Your calculaton!

Looks like it may have got partitioned with the wrong
geometry and that is what confused the imagers.

NO, I think it's just a calculation error
(understanding of what really needs to be accounted for).
Horst

 

[Horst Franke]

Okay, but why can't chkdsk handle this ?

Sorry Louis NO! CHKDSK seems to work OK,
but Your underlaying calculations were wrong.
And the other followons by Acronis, Paragon reflect on CHKDSK results!

So my outcome is: Your HDD has a problem which ist stated by CHKDSK!
You'll have to "repair" the disk before the other utilities will be able to
work correct! I know Your described behaviour from one of my disks, for
which I know that it has sector read (damage) problems.
Horst

 

[Folkert Rienstra]

Hostshit is back.

OK, Louis, I would treat this as a _real problem_!
Those things with felt back to 50 % are known by me for my HDD too!
This does not only look like logical problems but also like faulty sectors!

I also have such a device under Win98SE. On Start Up it sometimes take
TWO hours to recover from the faulty sectors (complete the Win98SE Boot)!


Yes that's same for me. NO, a Virus Scan has NOTHING to do with the
HDD consistency! Logical or sector failures are BESIDES any virus!

No problem, just delete the partition and recreate it.

But be aware, before that operation there should also be something
like a CHKDSK verification!
Even more, many partition creators implement a previous CHKDSK
in any form! So to be on a save way, DO it from Win before!
If there is a faulty area, then try to skip it by assigning it as UNUSED!

Ok, I don't know how effective the SMART report is.
Have too less experience (not enough years to verify) about SMART.

You have a problem with Your disk! You need to fix this before any
further actions!
Every partition change requires this check! Even any SW to create
a partition image!

How should a partition be saved when it is logically corrupted?
That means CHKDSK MUST run clean, before any of those actions!


Sure! It says there ARE problems with Your disk!
If You have a complete backup of that partition, then delete and recreate
it.

For MY specific case, I would have to mark a certain area as faulty, mark
it as UNUSABLE and move the new system partition AFTER that area.

Otherwise any Win98SE startup would need hours to recover that errors
and after reBoot it's again the same situation!
For me it's the boot-partition, which needs to be moved!

At the moment I tend to operate this HDD on a second PC in order
to perform the cleanup from there (HDD physically moved).
Horst

 

[Folkert Rienstra]

Horstshit is back.

Louis, so Your calculation was faulty!
Never mind, there is some overhead to be taken into account!
But last sector *535 is higher than *534!
But as this is very very near to the limit, it could also be a problem
of Acronis!

Have You already reported this to Acronis?
What Version of what special product have You used?
Horst

 

[Folkert Rienstra]

Horstshit is back.

And there is some more overhead (234.441.648 - 234.439.535) which
is about 2113 sectors (may be related to NTFS + some more)?

NO, the datasheets like above tells of
User Sectors Per Drive 234,441,648
which is above Your calculaton!



NO, I think it's just a calculation error
(understanding of what really needs to be accounted for).
Horst