I have a situation where I need to replace the directory hives on two data
disks (IOW, not current boot disks) from a currently bootable disk of
Win XP Pro SP3.

By selective disconnections from the IDE ribbon cable I observe repetitive
problems with the two data disks and I am told that the most likely
problem is a corrupted directory. The problem with both data disks when
used as boot disks:

1. POST is normal
2. BIOS verifies.
3. First windows screen displays for several seconds. (the black one with
a Windows logo.)
4. BSOD that appears and then disappear before any information can be
retrieved.
- return to restart.

(If i did not use all the right words, sorry.)

The current boot disk is showing signs of flakiness -- it had an uncalled
for in-session reboot yesterday. That tells me that its life is now on
loan. So i want the others to be ready to go when it fails.

All three disks have Windows XP pro installations and updates as follows:

hdd0 SP2, plus updates until SP3
hdd1 SP3 < -- currently booting disk
hdd2 SP2, plus updates until SP3

If Google search has details on this operation, I have not found them.

jim

jim wrote:
> I have a situation where I need to replace the directory hives on two data
> disks (IOW, not current boot disks) from a currently bootable disk of
> Win XP Pro SP3.
>
> By selective disconnections from the IDE ribbon cable I observe repetitive
> problems with the two data disks and I am told that the most likely
> problem is a corrupted directory. The problem with both data disks when
> used as boot disks:
>
> 1. POST is normal
> 2. BIOS verifies.
> 3. First windows screen displays for several seconds. (the black one with
> a Windows logo.)
> 4. BSOD that appears and then disappear before any information can be
> retrieved.
> - return to restart.
>
> (If i did not use all the right words, sorry.)
>
> The current boot disk is showing signs of flakiness -- it had an uncalled
> for in-session reboot yesterday. That tells me that its life is now on
> loan. So i want the others to be ready to go when it fails.
>
> All three disks have Windows XP pro installations and updates as follows:
>
> hdd0 SP2, plus updates until SP3
> hdd1 SP3 < -- currently booting disk
> hdd2 SP2, plus updates until SP3
>
> If Google search has details on this operation, I have not found them.
>
> jim

Have you run disk diagnostics on these three disks ?
For example, Seagate has Seatools, and Western Digital has diagnostics
available for download as well. If the disks pass, it could be
some portion of the file system is corrupted.

You can use HDTune "Health" tab, to examine the SMART data on the
three disks. You can also do a read-only bad block scan, using the
right-most tab in the interface.

http://www.hdtune.com/files/hdtune_255.exe

The CHKDSK utility when run, can attempt to repair file system damage.
Up to a point. Because it is a "repair in place" utility, it can
also ruin things. Which is where "backups" come in. You should
have full backups. And a fourth disk would provide a safe place
for them, if there isn't room on the other disks.

You can "repair" a file system, by copying the files off, reformatting
the partition, then copying the files back. The FAT (file allocation table)
or MFT (master file table) or whatever, would be refreshed, by starting
from scratch, and copying back the files.

What I would do (as a non-IT guy):

1) Review test results of the three disks. Did any fail ?
You want to back up the failed disk right away.
If the damage is severe, you do a sector by sector backup.
If the damage is light, a file by file backup is sufficient,
since your expectation is, all the files are intact. The sector
by sector is used, when you fear losing the whole thing.

2) Do a bad block scan, to check whether any sectors are unreadable.
If there are unreadable sectors, which are not cordoned off in
a bad cluster, then things associated with those sectors would be
broken. If the bad block scan is clean, it means maintenance
will be easier.

In this example, a badly damaged disk is copied off sector by sector.
Later, you'd copy this image back, to the new disk you bought. And
then, do further repair work, in the knowledge that any further writes
won't get corrupted.

http://www.cgsecurity.org/wiki/Damaged_Hard_Disk

# first, grab most of the error-free areas in a hurry:
./ddrescue -n /dev/old_disk /dev/new_disk rescued.log
# then try to recover as much of the dicy areas as possible:
./ddrescue -r 1 /dev/old_disk /dev/new_disk rescued.log

3) If all disks were healthy, and this was only a file system level
issue, I'd

a) boot hdd2 SP2
b) Create a partition the size of hdd1 SP3, with matching file system type,
in spare area on hdd2 or hdd0.
c) Copy the contents of hdd1 SP3 to the new partition just created.

http://en.wikipedia.org/wiki/Robocopy
# Download site
http://technet.microsoft.com/en-us/m...spotlight.aspx

If I wanted to copy all the files off a partition with letter "J:" to
a *completely empty* partition "E:", I'd do it like this.

robocopy J:\ E:\ /mir /copy:datso /dcopy:t /r:3 /w:2 /zb /np /tee /v /log:robocopy_J_to_E.log

d) Once the copy is complete (could take 20 minutes, so relax), next
I'd go into Disk Management and reformat the partition I just copied off.
If the source partition was J:, I'd be reformatting it, to the same
file system standard as it had before. If it was an oversized FAT32
partition, and Windows refused to format it, I'd club it into submission
with the fat32formatter, as in fat32formatter.exe J:

http://www.ridgecrop.demon.co.uk/dow...at32format.zip

The Disk Management NTFS format capability, by comparision, doesn't have
issued, so no third-party would be needed in that case.

e) Now, I can put the files back on the clean J: like this.

robocopy E:\ J:\ /mir /copy:datso /dcopy:t /r:3 /w:2 /zb /np /tee /v /log:robocopy_E_to_J.log

f) Copying off the files like that, does not copy the "partition boot sector".
To fix that, you would shut down, disconnect the good disks HDD0 and HDD2,
(to avoid confusion), boot with the WinXP installer CD, so you could
access the Recovery Console. And run a "fixboot c:" from there. That
will put back the partition boot sector on hdd1's new WinXP partition.

g) (You could reconnect the other two disks at this point if you want.)

h) For extra points, you could use "VolumeID" to put back the original
volume ID that HDD1 had. But perhaps this is too much nit-picking.
I get the VolumeID with Everest (free) Home Edition, then restore
it later.

http://technet.microsoft.com/en-us/s...rnals/bb897436

i) At this point, the idea would be, the HDD1 WinXP partition has a
fresh file system, with minty fresh FAT or MFT, and all the files
copied back. This does not preclude damage happening along the
way, and a disaster awaiting when you boot the system with that disk.
Now, you adjust the BIOS boot order, so that HDD1 is again selected
to boot.

j) If any step fails along the way, you'd have that copy on E: to
save you. And you wouldn't erase J:, until you'd satisfied yourself
the copy on E: was complete.

Anyway, I doubt you're going to do all that, but there are tools
around, command line and otherwise, that allow repair attempts.

In the above, I've made no mention of "directory specific" repairs,
or "copying a registry file" or the like. Because at this point,
I don't really know what's broken. There are undoubtedly other
issues there.

And if you want to run CHKDSK on that drive, you'd boot from
HDD0 or HDD2 and run it from there, on the now "non-busy" HDD1
WinXP partition. Otherwise, CHKDSK on a busy partition, can only
run during the boot process. Since you have multiple OS partitions,
you can boot one of those, and then CHKDSK can run at an arbitrary
time. But I would not be using CHKDSK at all, until I had my
backup in place. A backup done with "ddrescue" method, if the
disk is really in a mess. Or, using some other $39.95 type utility
(Acronis) if it is in relatively good shape.

Summary:

1) Do backups, before using "in-place" repair tools such as CHKDSK.
Do backups, "before it is too late". If the drive is "clicking",
backup now! If you hear clicking, you might also want to backup
without powering off or rebooting, in case the drive is going to
die in the next ten minutes.
2) You can repair a file system, by copying the files off while the
partition is not busy.
3) There are a ton of half-baked utilities around, for fixing this
and that.

Paul

jim wrote:
> I have a situation where I need to replace the directory hives on two data
> disks (IOW, not current boot disks) from a currently bootable disk of
> Win XP Pro SP3.
>
> By selective disconnections from the IDE ribbon cable I observe repetitive
> problems with the two data disks and I am told that the most likely
> problem is a corrupted directory. The problem with both data disks when
> used as boot disks:
>
> 1. POST is normal
> 2. BIOS verifies.
> 3. First windows screen displays for several seconds. (the black one with
> a Windows logo.)
> 4. BSOD that appears and then disappear before any information can be
> retrieved.
> - return to restart.
>
> (If i did not use all the right words, sorry.)
>
> The current boot disk is showing signs of flakiness -- it had an uncalled
> for in-session reboot yesterday. That tells me that its life is now on
> loan. So i want the others to be ready to go when it fails.
>
> All three disks have Windows XP pro installations and updates as follows:
>
> hdd0 SP2, plus updates until SP3
> hdd1 SP3 < -- currently booting disk
> hdd2 SP2, plus updates until SP3
>
> If Google search has details on this operation, I have not found them.
>
> jim


"BSOD that appears and then disappear before any information can be retrieved"

You can change the behavior of WinXP, by using the "automatic restart" tick
box. But of course, it's too late to do that now, if you cannot get into
Windows.

You see, this is a stupid default setting (box ticked), as it prevents
easily copying the BSOD message. Having the box ticked might be appropriate
for a "kiosk" machine, but not for a desktop. When the box is unticked,
the BSOD will stand still until you've copied it, and restarted.

(Each OS will have some version of this...)

http://cdn.nirmaltv.com/images/disab...ticrestart.png

You could always try Windows Safe Mode (press F8), and perhaps then you
could get to the control panel and change that setting.

Paul

On Mon, 07 May 2012 12:15:30 -0400, in
microsoft.public.windowsxp.general, Paul <(E-Mail Removed)>, wrote

>jim wrote:
>> I have a situation where I need to replace the directory hives on two data
>> disks (IOW, not current boot disks) from a currently bootable disk of
>> Win XP Pro SP3.
>>
>> By selective disconnections from the IDE ribbon cable I observe repetitive
>> problems with the two data disks and I am told that the most likely
>> problem is a corrupted directory. The problem with both data disks when
>> used as boot disks:
>>
>> 1. POST is normal
>> 2. BIOS verifies.
>> 3. First windows screen displays for several seconds. (the black one with
>> a Windows logo.)
>> 4. BSOD that appears and then disappear before any information can be
>> retrieved.
>> - return to restart.
>>
>> (If i did not use all the right words, sorry.)
>>
>> The current boot disk is showing signs of flakiness -- it had an uncalled
>> for in-session reboot yesterday. That tells me that its life is now on
>> loan. So i want the others to be ready to go when it fails.
>>
>> All three disks have Windows XP pro installations and updates as follows:
>>
>> hdd0 SP2, plus updates until SP3
>> hdd1 SP3 < -- currently booting disk
>> hdd2 SP2, plus updates until SP3
>>
>> If Google search has details on this operation, I have not found them.
>>
>> jim
>
>Have you run disk diagnostics on these three disks ?

Yes. Numerous Chkdsk /F and Chkdsk /R, in-session for non-boot disks and
at boot for boot disk. A question of a while back was what would happen
if you scheduled 2 for the same 'next' boot -- answer: it stacks the
commands, i.e., they execute one at a time.

1 of the disks is currently disconnected, so what follows applies to what
above is listed as hdd1 and hdd2.

S.M.A.R.T, extended test -- hdd2: no error, On line hddstatus tool
http://www.hddstatus.com/hdrepanalysis.php hdd1: not so good -- 19 pending
writes and 19 uncorrectable sectors.
(Same hddstatus tool reported all "normal values for hdd2)

I will look into the other steps you mention.

>For example, Seagate has Seatools, and Western Digital has diagnostics
>available for download as well. If the disks pass, it could be
>some portion of the file system is corrupted.
>
>You can use HDTune "Health" tab, to examine the SMART data on the
>three disks. You can also do a read-only bad block scan, using the
>right-most tab in the interface.
>
>http://www.hdtune.com/files/hdtune_255.exe
>
>The CHKDSK utility when run, can attempt to repair file system damage.
>Up to a point. Because it is a "repair in place" utility, it can
>also ruin things. Which is where "backups" come in. You should
>have full backups. And a fourth disk would provide a safe place
>for them, if there isn't room on the other disks.
>
>You can "repair" a file system, by copying the files off, reformatting
>the partition, then copying the files back. The FAT (file allocation table)
>or MFT (master file table) or whatever, would be refreshed, by starting
>from scratch, and copying back the files.
>
>What I would do (as a non-IT guy):
>
>1) Review test results of the three disks. Did any fail ?
> You want to back up the failed disk right away.
> If the damage is severe, you do a sector by sector backup.
> If the damage is light, a file by file backup is sufficient,
> since your expectation is, all the files are intact. The sector
> by sector is used, when you fear losing the whole thing.
>
>2) Do a bad block scan, to check whether any sectors are unreadable.
> If there are unreadable sectors, which are not cordoned off in
> a bad cluster, then things associated with those sectors would be
> broken. If the bad block scan is clean, it means maintenance
> will be easier.
>
> In this example, a badly damaged disk is copied off sector by sector.
> Later, you'd copy this image back, to the new disk you bought. And
> then, do further repair work, in the knowledge that any further writes
> won't get corrupted.
>
> http://www.cgsecurity.org/wiki/Damaged_Hard_Disk
>
> # first, grab most of the error-free areas in a hurry:
> ./ddrescue -n /dev/old_disk /dev/new_disk rescued.log
> # then try to recover as much of the dicy areas as possible:
> ./ddrescue -r 1 /dev/old_disk /dev/new_disk rescued.log
>
>3) If all disks were healthy, and this was only a file system level
> issue, I'd
>
> a) boot hdd2 SP2
> b) Create a partition the size of hdd1 SP3, with matching file system type,
> in spare area on hdd2 or hdd0.
> c) Copy the contents of hdd1 SP3 to the new partition just created.
>
> http://en.wikipedia.org/wiki/Robocopy
> # Download site
> http://technet.microsoft.com/en-us/m...spotlight.aspx
>
> If I wanted to copy all the files off a partition with letter "J:" to
> a *completely empty* partition "E:", I'd do it like this.
>
> robocopy J:\ E:\ /mir /copy:datso /dcopy:t /r:3 /w:2 /zb /np /tee /v /log:robocopy_J_to_E.log
>
> d) Once the copy is complete (could take 20 minutes, so relax), next
> I'd go into Disk Management and reformat the partition I just copied off.
> If the source partition was J:, I'd be reformatting it, to the same
> file system standard as it had before. If it was an oversized FAT32
> partition, and Windows refused to format it, I'd club it into submission
> with the fat32formatter, as in fat32formatter.exe J:
>
> http://www.ridgecrop.demon.co.uk/dow...at32format.zip
>
> The Disk Management NTFS format capability, by comparision, doesn't have
> issued, so no third-party would be needed in that case.
>
> e) Now, I can put the files back on the clean J: like this.
>
> robocopy E:\ J:\ /mir /copy:datso /dcopy:t /r:3 /w:2 /zb /np /tee /v /log:robocopy_E_to_J.log
>
> f) Copying off the files like that, does not copy the "partition boot sector".
> To fix that, you would shut down, disconnect the good disks HDD0 and HDD2,
> (to avoid confusion), boot with the WinXP installer CD, so you could
> access the Recovery Console. And run a "fixboot c:" from there. That
> will put back the partition boot sector on hdd1's new WinXP partition.
>
> g) (You could reconnect the other two disks at this point if you want.)
>
> h) For extra points, you could use "VolumeID" to put back the original
> volume ID that HDD1 had. But perhaps this is too much nit-picking.
> I get the VolumeID with Everest (free) Home Edition, then restore
> it later.
>
> http://technet.microsoft.com/en-us/s...rnals/bb897436
>
> i) At this point, the idea would be, the HDD1 WinXP partition has a
> fresh file system, with minty fresh FAT or MFT, and all the files
> copied back. This does not preclude damage happening along the
> way, and a disaster awaiting when you boot the system with that disk.
> Now, you adjust the BIOS boot order, so that HDD1 is again selected
> to boot.
>
> j) If any step fails along the way, you'd have that copy on E: to
> save you. And you wouldn't erase J:, until you'd satisfied yourself
> the copy on E: was complete.
>
>Anyway, I doubt you're going to do all that, but there are tools
>around, command line and otherwise, that allow repair attempts.
>
>In the above, I've made no mention of "directory specific" repairs,
>or "copying a registry file" or the like. Because at this point,
>I don't really know what's broken. There are undoubtedly other
>issues there.
>
>And if you want to run CHKDSK on that drive, you'd boot from
>HDD0 or HDD2 and run it from there, on the now "non-busy" HDD1
>WinXP partition. Otherwise, CHKDSK on a busy partition, can only
>run during the boot process. Since you have multiple OS partitions,
>you can boot one of those, and then CHKDSK can run at an arbitrary
>time. But I would not be using CHKDSK at all, until I had my
>backup in place. A backup done with "ddrescue" method, if the
>disk is really in a mess. Or, using some other $39.95 type utility
>(Acronis) if it is in relatively good shape.
>
>Summary:
>
>1) Do backups, before using "in-place" repair tools such as CHKDSK.
> Do backups, "before it is too late". If the drive is "clicking",
> backup now! If you hear clicking, you might also want to backup
> without powering off or rebooting, in case the drive is going to
> die in the next ten minutes.
>2) You can repair a file system, by copying the files off while the
> partition is not busy.
>3) There are a ton of half-baked utilities around, for fixing this
> and that.
>
> Paul

"jim" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
>I have a situation where I need to replace the directory hives on two
>data
> disks (IOW, not current boot disks) from a currently bootable disk of
> Win XP Pro SP3.
>
> By selective disconnections from the IDE ribbon cable I observe
> repetitive
> problems with the two data disks and I am told that the most likely
> problem is a corrupted directory. The problem with both data disks
> when
> used as boot disks:
>
> 1. POST is normal
> 2. BIOS verifies.
> 3. First windows screen displays for several seconds. (the black one
> with
> a Windows logo.)
> 4. BSOD that appears and then disappear before any information can be
> retrieved.
> - return to restart.
>
> (If i did not use all the right words, sorry.)
>
> The current boot disk is showing signs of flakiness -- it had an
> uncalled
> for in-session reboot yesterday. That tells me that its life is now
> on
> loan. So i want the others to be ready to go when it fails.
>
> All three disks have Windows XP pro installations and updates as
> follows:
>
> hdd0 SP2, plus updates until SP3
> hdd1 SP3 < -- currently booting disk
> hdd2 SP2, plus updates until SP3
>
> If Google search has details on this operation, I have not found them.

If you repeatedly tap F8 at startup, you should get to the boot menu.
There should be an option on the boot menu to Disable Automatic
Restart.... select that entry and press Enter. That should allow you to
see the BSOD screen.

If the option is not listed, select Safe Mode from the boot menu, and
see if that allows you to start Windows so you can disable automatic
restart from the System Control Panel.
--
Glen Ventura
MS MVP Oct. 2002 - Sept. 2009
CompTIA A+

On Mon, 07 May 2012 12:15:30 -0400, in
microsoft.public.windowsxp.general, Paul <(E-Mail Removed)>, wrote

>The CHKDSK utility when run, can attempt to repair file system damage.
>Up to a point. Because it is a "repair in place" utility, it can
>also ruin things.


Cold startup was normal 5/8/12AM.

I set hdd1 (seagate 250, current boot disk) to run CHKDSK /F next startup.

Did a warm start, CHKDSK ran in ~2 minutes. No errors found, then
continuation of startup and BSOD and then directly to automatic restart
which was normal.

I found this on web to disable automatic restart:
"Right Click on “My Computer” and goto “Properties”. Then goto the
“Advanced” tab and under the headline “Startup and Recovery”, press the
“Settings” button. Untick the “Automatically Restart” box and press Ok."

Did it.

Will look into Seatools utilities in a few hours when i wake up decently.

This disk had all S.M.A.R.T parameters normal EXCEPT "pending rewrites"
and "uncorrectable sector errors". BTW, it has a 10 year warranty and i
am 4 years in -- not that i have a clue as to how/where to redeem
warranty, it just gives me an idea of expected life span.

jim

On Mon, 7 May 2012 11:42:25 -0500, in microsoft.public.windowsxp.general,
rb <(E-Mail Removed)>, wrote

>
>jim;1339575 Wrote:
>> I have a situation where I need to replace the directory hives on two
>> data
>> disks (IOW, not current boot disks) from a currently bootable disk of
>> Win XP Pro SP3.
>>
>> By selective disconnections from the IDE ribbon cable I observe
>> repetitive
>> problems with the two data disks and I am told that the most likely
>> problem is a corrupted directory. The problem with both data disks when
>> used as boot disks:
>>
>> 1. POST is normal
>> 2. BIOS verifies.
>> 3. First windows screen displays for several seconds. (the black one
>> with
>> a Windows logo.)
>> 4. BSOD that appears and then disappear before any information can be
>> retrieved.
>> - return to restart.
>>
>> (If i did not use all the right words, sorry.)
>>
>> The current boot disk is showing signs of flakiness -- it had an
>> uncalled
>> for in-session reboot yesterday. That tells me that its life is now on
>> loan. So i want the others to be ready to go when it fails.
>>
>> All three disks have Windows XP pro installations and updates as
>> follows:
>>
>> hdd0 SP2, plus updates until SP3
>> hdd1 SP3 < -- currently booting disk
>> hdd2 SP2, plus updates until SP3
>>
>> If Google search has details on this operation, I have not found them.
>>
>> jim
>
>
>
>Try doing a 'Repair' install of XP....
>
>Boot from your XP CD.
>XP Setup will begin loading. You will be presented with the "Windows
>Setup" screen.
>Select 'To set up Windows XP now, press ENTER'. *DO NOT* select Recovery
>Console. (It will appear as if you are doing a 'fresh install', but you
>will be given another 'Repair' option.)

I have been at this "t will appear as if you are doing a 'fresh install' "
point several times in the past and it has always been 'scary'. :-) I
always crash out of it, but with your information, will take if further
next time.

>Accept the license agreement by pressing F8 key.
>Windows Setup will look for existing XP installation(s).
>
>Note: If no installations are found, then you will not be given the
>option to repair. This may happen if the data or partition on your drive
>is too corrupted.
>
>If it finds an existing XP installation, you will be presented with the
>option to repair the existing XP installation, or install a fresh copy
>of XP.
>Select the option 'To repair the selected XP installation' by pressing
>'R'.
>
>Note: If you install a fresh copy at this point, *all data on that
>partition will be lost!*
>
>XP will appear to be installing itself for the first time, but it will
>retain all of your data and settings. Just follow the prompts, and have
>your XP CD Key ready if needed.
>
Thanks,

jim

On Mon, 7 May 2012 20:27:00 -0400, in microsoft.public.windowsxp.general,
"glee" <(E-Mail Removed)>, wrote

>If you repeatedly tap F8 at startup, you should get to the boot menu.
>There should be an option on the boot menu to Disable Automatic
>Restart.... select that entry and press Enter. That should allow you to
>see the BSOD screen.

Thanks, Glee, it appears that i got it earlier this AM from a normal
windows startup.

I believe it is a register dump -- which part of it is particularly useful
-- or is it just "all of it"?

jim

On Mon, 07 May 2012 12:20:45 -0400, in
microsoft.public.windowsxp.general, Paul <(E-Mail Removed)>, wrote

>jim wrote:
>> I have a situation where I need to replace the directory hives on two data
>> disks (IOW, not current boot disks) from a currently bootable disk of
>> Win XP Pro SP3.
etc....

To everyone, this may be helpful
http://dl.dropbox.com/u/14875021/compmanage5_8.jpg

jim

jim wrote:
> On Mon, 07 May 2012 12:20:45 -0400, in
> microsoft.public.windowsxp.general, Paul <(E-Mail Removed)>, wrote
>
>> jim wrote:
>>> I have a situation where I need to replace the directory hives on two data
>>> disks (IOW, not current boot disks) from a currently bootable disk of
>>> Win XP Pro SP3.
> etc....
>
> To everyone, this may be helpful
> http://dl.dropbox.com/u/14875021/compmanage5_8.jpg
>
> jim
>

You can actually open one of those Event errors, for a look.

This is the entry for Event ID 11, which you can compare against
the text actually in the error on your machine.

http://www.eventid.net/display.asp?e...e=Disk&phase=1

"In certain conditions, this error message is related to the S.M.A.R.T attribute 187
(Reported uncorrectable errors). When a physical sector of the disk is read by an
application (e.g. a SMART test tool or a simple copy command) and a sector is
"unstable", meaning that the drive electronics (hardware ECC) is unable to
correctly read the sector, then Windows logs Event ID 11."

You can use HDTune "Health" tab to read the current statistics.
If you convert 187 decimal to hex, that ix 0xBB, and when I display
that on my disk, it's listed as "unknown attribute".

What I use as a health indicator, is Reallocated Sector Count (effectively
past history), and Current Pending Sector (which is a count of sectors
needing to be processed for instability). The Current Pending might grow
as a function of the Event ID 11. The Current Pending gets reduced, if
on a write operation, the controller is able to decide what to do with
the sector. A sector either tests as being stable, in which case it
doesn't need to be substituted, and Current Pending can be decremented.
Or, a sector tests as being corrupt after a write, in which case the
controller uses a spare sector and makes an entry in a substitution table.
Again, Current Pending is decremented after the processing of a sector.

If I get non-zero data columns for those two entries, I consider replacing
the drive. And that's where my average two drive purchases per year come
from - a conservative disk operating policy.

Paul