Why wont XP Defragger place all data together?

[casey.o]

The defragger in Win98 (actually an import from WinME), dont leave any
gaps in the data after it completes. But the one in XP leaves all kinds
of gaps. I've run it 6 times in a row.... Why cant it get all the data
put to the front of the drive?

One other thing, the XP defrag is a lot slower than the one from WinME?
It reminds me of the old one from Win3.1..... Slowwwwwwwww.......

I like the look of the one in Win98 better too, because you can see it
working, but that's not real important....

 

[casey.o]

IIRC, the disk defragger in Windows 98 was licensed from Executive
Software International, Inc. It spawned Diskeeper, Inc., and marketed
its own version under the name of Diskeeper. For some reason or the
other, Microsoft was compelled to dissociate itself from the makers of
Diskeeper.

Today, Executive Software International, Inc., or Diskeeper, Inc., is
known as Condusive Technologies.

Interesting.....
I seem to recall hearing something about this back in the Win3.x days...

You'd think MS would have just bought the company, they can afford it.
I'm not too pleased with the XP defragger, when compared top the Win98
version, especially the speed of it. (I'm actually referring to the one
from WinME, which anyone who knows anything about Win98, put the ME one
on their computer). It was the fastest and the last time it was used.
Win2k has the same as XP.

I always liked watchin those boxes move around too....
There's nothing to watch in XP or 2K.
I dont know what the newer versions of Wondows use.

 

[micky]

Interesting.....
I seem to recall hearing something about this back in the Win3.x days...

You'd think MS would have just bought the company, they can afford it.
I'm not too pleased with the XP defragger, when compared top the Win98
version, especially the speed of it. (I'm actually referring to the one
from WinME, which anyone who knows anything about Win98, put the ME one
on their computer).

You should really just call it the ME defragger then. I had no idea
what you meant in your OP.

It's certainly absurd to say that "anyone who knows anything about
Win98, put the ME one on their computer".

It was the fastest and the last time it was used.
Win2k has the same as XP.

I always liked watchin those boxes move around too....

Yes. Me too. I also had Norton defragger and it would put many of
the files at the other end of the disk, while win98 put them all near
the start. So if one were perverse, he could alternate between 98 and
norton and there was always a lot of defragging-work to do.

 

[Rasta Robert]

The defragger in Win98 (actually an import from WinME), dont leave any
gaps in the data after it completes. But the one in XP leaves all kinds
of gaps. I've run it 6 times in a row.... Why cant it get all the data
put to the front of the drive?

One other thing, the XP defrag is a lot slower than the one from WinME?
It reminds me of the old one from Win3.1..... Slowwwwwwwww.......

I like the look of the one in Win98 better too, because you can see it
working, but that's not real important....

There are some third party defraggers that give more control over
defragmenting, and give a visual representation while working.
One I like a lot is the portable version of Defraggler by Piriform,
who also make Ccleaner.
Portable and Slim (no toolbar) versions can be downloaded from
<http://www.piriform.com/defraggler/builds>

 

[Henry]

The defragger in Win98 (actually an import from WinME), dont leave any
gaps in the data after it completes. But the one in XP leaves all kinds
of gaps. I've run it 6 times in a row.... Why cant it get all the data
put to the front of the drive?

One other thing, the XP defrag is a lot slower than the one from WinME?
It reminds me of the old one from Win3.1..... Slowwwwwwwww.......

I like the look of the one in Win98 better too, because you can see it
working, but that's not real important....

Use Auslogics. Works like old defrager and is faster than the one with
WinXP. Auslogics also has a registry cleaner and both programs are free.
Google it.

 

[philo]

IIRC, the disk defragger in Windows 98 was licensed from Executive
Software International, Inc. It spawned Diskeeper, Inc., and marketed
its own version under the name of Diskeeper. For some reason or the
other, Microsoft was compelled to dissociate itself from the makers of
Diskeeper.

Today, Executive Software International, Inc., or Diskeeper, Inc., is
known as Condusive Technologies.

GR

Also, Win98 is installed on a Fat32 partition

and XP is more than likely on NTFS


NTFS will have an MFT (Master file table) that will not get defragged,
but that will have little if any effect on performance.


Bottom line: XP defrag is good enough

 

[Zo]

Henry formulated the question :

Use Auslogics. Works like old defrager and is faster than the one with
WinXP. Auslogics also has a registry cleaner and both programs are free.
Google it.

+1 on this one. I use the portable version.

--
Zo

"The trouble with having an open mind, of course, is that people
will insist on coming along and trying to put things in it."
-- Terry Pratchett (Diggers)

 

[Stef]

The defragger in Win98 (actually an import from WinME), dont leave any
gaps in the data after it completes. But the one in XP leaves all kinds
of gaps. I've run it 6 times in a row.... Why cant it get all the data
put to the front of the drive?

Don't worry about it. This is normal. It's one of the differences
between how FAT32 and NTFS filesystems work. In NTFS, it actually
reduces file fragmentation. It also reduces overall hard drive ware and
tear by distributing writes over the entire hard drive instead of
concentrating on one particular region as a defragged FAT32 drive are
prone to do.

One other thing, the XP defrag is a lot slower than the one from WinME?
It reminds me of the old one from Win3.1..... Slowwwwwwwww.......

Nature of the beast with NTFS defrags. But with NTFS, you shouldn't
need to defrag all that often. So, it balances out.

I like the look of the one in Win98 better too, because you can see it
working, but that's not real important....

Don't get caught up in the pretty graphics. The important thing is
improving read/write efficiency whether you're using FAT32 or NTFS.
FYI: NTFS is faster and more efficient than FAT32.

Stef

 

[casey.o]

Don't worry about it. This is normal. It's one of the differences
between how FAT32 and NTFS filesystems work. In NTFS, it actually
reduces file fragmentation. It also reduces overall hard drive ware and
tear by distributing writes over the entire hard drive instead of
concentrating on one particular region as a defragged FAT32 drive are
prone to do.


Nature of the beast with NTFS defrags. But with NTFS, you shouldn't
need to defrag all that often. So, it balances out.


Don't get caught up in the pretty graphics. The important thing is
improving read/write efficiency whether you're using FAT32 or NTFS.
FYI: NTFS is faster and more efficient than FAT32.

Stef

The drive in this old Compaq is formatted to Fat32. That's how I got
it. The drive in my laptop, is NTFS. Both defrag the same. Same slow
speed, same appearance. Actually the secondary boot on my Win98
computer is Win2000. That looks and acts just like the one in XP. But
Win2000 is said to be the earliest version of XP anyhow, which makes
sense, just like Win95 was where Win98 began.

While Fat32 is supposed to be more efficient, I personally prefer the
Fat32, simply because if the system was to crash, I can still access the
data from booting to Dos. I wonder if it would make the most sense to
set the C: drive, to NTFS, keep that a small partition, and use it
mostly just for the OS and programs. Then use Fat32 partitions and/or a
second physicla HD for data storage, and have that set to Fat32???

 

[Paul]

The defragger in Win98 (actually an import from WinME), dont leave any
gaps in the data after it completes. But the one in XP leaves all kinds
of gaps. I've run it 6 times in a row.... Why cant it get all the data
put to the front of the drive?

One other thing, the XP defrag is a lot slower than the one from WinME?
It reminds me of the old one from Win3.1..... Slowwwwwwwww.......

I like the look of the one in Win98 better too, because you can see it
working, but that's not real important....

There are two properties, defragmenting and optimizing.

Defragmenting, takes a file and locates all the clusters
so they are contiguous. This helps reduce the time to read
the file later. If the file is fragmented, more head movement
on the disk drive, is required to find all the pieces.

Optimizing, is the moving or sorting of files in a particular way.
"Slam to the left", is a popular optimization strategy for
partitions, reducing the average size of head movements as
long as the partition is only partially full.

And example of a "pure defragmenter" is Sysinternals "contig.exe".
It attempts to place all the pieces of a file next to one
another, but it doesn't care what part of the partition
gets the file. The file could end up over on the right.
The contig.exe does no optimization at all, and that's why it
is the purest defragmenter. Most other tools you've used,
will include optimization (even if it's relatively pitiful).

The WinXP defragmenter, has two kinds of optimization. It may rearrange
things like ".pf" prefetch files, when the system defragmenter is used.
And, it also throws the files to the left. For files which are
"movable".

The WinXP one has a display. The top bar shows the "before", and the bottom
bar shows the "after". And one of the colors identifies "unmovable"
content. The unmovable things might be metafiles, or things
the OS is currently using (pagefile.sys).

http://i59.tinypic.com/2rqyr1s.gif

Windows has an API for data movement, used by the defragmenters.
It was initially added to the OS, after Microsoft worked with one
of the defragmenter companies. The data movement is noteworthy,
in that only tiny movements are supported, like reading and writing
of 4KB and 64KB chunks. This is one of the reasons defragmentation
can be slow, because inefficient commands are given to the hard drive.
The result is data transfer rates in the 1 MB/sec to 3 MB/sec range.
The purpose of doing the small movements, is to make defragmentation
"power safe". If the power goes off, the idea is, the partition survives.
Even though data is in transit. If really large data movements were
attempted, that property might no longer be true.

*******

Tools:

JKDefrag, is a free defragmenter, with various optimization policies.
It uses an exotic policy, of keeping "large" files in a group
on the right hand side of the partition. It also has command line
options, to "slam to the left", where it doesn't worry about
file fragmentation while doing so. This is convenient if you're
in a hurry to do a partition shrink, and don't care what the
resulting (small) partition looks like. That's useful for
data partitions you don't plan on growing, where the files will
be left to stagnate on the partition.

Raxco PerfectDisk is a commercial application. A trial is available.
Raxco can move more metadata files, than the Windows defragmenters
would agree to. And it does that stuff, without a reboot or anything.
It's pretty amazing.

Partition Manager applications, they "slam to the left" as well, but
I doubt they spend any time "making the data pretty". So a Partition
Manager can move stuff, but the level of fragmentation could
actually increase while it is doing it. That's because when you
buy a Partition Manager, its sole job is just to resize partitions
as quickly as possible. No display is offered, to show the level
of fragmentation.

http://en.wikipedia.org/wiki/Ntfs

"Resizing

Starting with Windows Vista, Microsoft added the built-in ability
to shrink or expand a partition, but this capability is limited
because it will not relocate page file fragments or files that
have been marked as unmovable. So shrinking will often require
relocating or disabling any page file, the index of Windows Search,
and any Shadow Copy used by System Restore. Various third-party
tools are capable of resizing NTFS partitions."

*******

If you're a "neat freak", you can spend time moving the Windows
Search index, flushing all the System Restore points, disable VSS,
disable the pagefile.sys, and prepare the partition for defragmentation.
But doing all of that, you may still run into a few tiny things
that won't move. Which of course, would drive you crazy

Some of the patterns you see on the disk, exist because the file
system is reserving space for the $MFT to grow. And then the
defragmentation you attempted, may not successfully slam everything
to the left.

*******

Before showing this section, I have to make the usual disclaimers:

1) Always back up, before doing intensive data movements.
CHKDSK, defragmentation, partition management, may move lots
of data around. And the programs can severely mess up your
disk, once in a while.

2) When using a new and untested program, the risks are much
higher on the first run. You don't really know what's going to
happen. Only after a lot of usage, will you begin to trust
the new tool you're testing.

So before trying these things, you'd want a backup in hand. Just in case.

http://www.kessels.com/JkDefrag/ <--- Jerome Kessels free defrag, with source
Good for experiments

http://www.mydefrag.com/Manual-DownloadAndInstall.html <--- New version by JK, no src

Example of a command line invocation (using the first package)

JkDefrag.exe -a 5 -d 2 -q c:

-a = optimize mode
-d = debug level <--- places details in a .log file
-q = quit_on_finish

-a N
1 = Analyze, do not defragment and do not optimize.
2 = Defragment only, do not optimize.
3 = Defragment and fast optimize [recommended]. <--- default
5 = Force together.
6 = Move to end of disk.
7 = Optimize by sorting all files by name (folder + filename).
8 = Optimize by sorting all files by size (smallest first).
9 = Optimize by sorting all files by last access (newest first).
10 = Optimize by sorting all files by last change (oldest first).
11 = Optimize by sorting all files by creation time (oldest first).

-d N
0 = Fatal errors.
1 = Warning messages [default].
2 = General progress messages.
3 = Detailed progress messages.
4 = Detailed file information.
5 = Detailed gap-filling messages.
6 = Detailed gap-finding messages.

I would make a backup first, then play with the program,
to see what it can do for you.

I just tried the "-a 5" option on C: in a WinXP test VM,
and it does "slam to the left". It cannot move the "green bar"
of pagefile.sys, and didn't appear to try in this case.
The version of JKDefrag I used was 3.36.

HTH,
Paul

 

[micky]

Henry formulated the question :

+1 on this one. I use the portable version.

This thread reminded me that I wanted to defrag, but when it was 94%
done with the built-in XP defrag analysiing, it found some file that
had some problem with it. I clicked OK on the message box, thinking
the program woudl skip this file and keep going. Instead it stopped and
I have to start from the beginning. I don't even need the file.

Do you think Auslogics or some other defrag will handle a bad file
without stopping the program?

 

[Zo]

micky formulated the question :

This thread reminded me that I wanted to defrag, but when it was 94%
done with the built-in XP defrag analysiing, it found some file that
had some problem with it. I clicked OK on the message box, thinking
the program woudl skip this file and keep going. Instead it stopped and
I have to start from the beginning. I don't even need the file.

Do you think Auslogics or some other defrag will handle a bad file
without stopping the program?

The only person who will know the answer to that is you. Give it a try
and let us know how it turns out.

Here's the link http://www.auslogics.com/en/software/disk-defrag/

Portable version is near the bottom of the page on the right.

 

[Paul]

This thread reminded me that I wanted to defrag, but when it was 94%
done with the built-in XP defrag analysiing, it found some file that
had some problem with it. I clicked OK on the message box, thinking
the program woudl skip this file and keep going. Instead it stopped and
I have to start from the beginning. I don't even need the file.

Do you think Auslogics or some other defrag will handle a bad file
without stopping the program?

Um, don't do that!

Any time data-intensive things are run, even the tools
themselves will use CHKDSK to determine whether it's "safe"
or not. For example, Partition Managers which are about to
resize things, check whether the file system is intact
first, before doing anything dangerous.

The same logic should apply to defragmenters. You shouldn't
run a defragmenter, unless the file system is solid.

I would say you have a much more serious problem than
making the partition pretty, and should attend to that error first.
If you think deleting the file is going to fix it, go right
ahead. My guess is, it'll still be broken when you're done.
Run CHKDSK again and see.

While you're at it, you might check your SMART statistics,
and see if the hardware is good. It could be, that the
error detected, is related to actual bad hardware. And
you don't want to use either CHKDSK or defragmenters
on flaky hardware. Install a new disk, before letting
those kinds of tools loose.

(The ones in red boxes, should be zero.)

http://oi62.tinypic.com/2wghquh.jpg

Since this has *never* happened to me here, I'd be at least
a bit curious as to the root cause. You might actually
be in serious trouble (a day away from losing a hard drive).

Paul

 

[Stef]

The drive in this old Compaq is formatted to Fat32. That's how I got
it. The drive in my laptop, is NTFS. Both defrag the same. Same slow
speed, same appearance. Actually the secondary boot on my Win98
computer is Win2000. That looks and acts just like the one in XP. But
Win2000 is said to be the earliest version of XP anyhow, which makes
sense, just like Win95 was where Win98 began.

What "stanardard" are you using for defragging speed comparison? What
are the specs of these "slow" systems?

Defragging speed can be affected by a number of factors, but the biggest
two are a very full hard drive leaving too little space to "buffer"
files during the defrag, and lack of free RAM. Check. If your drive is
over 90% full, you're going to have problems. Slow defragging being
one.

Use diskcleaner or something similar to clear old and unneeded files off
your hard drive before defragging. Remove unneeded background processes
and startup apps to free up RAM, particularly if you are short on RAM.

Scan for trojans, viruses and malware. They can greatly affect system
performance.

While Fat32 is supposed to be more efficient, I personally prefer the
Fat32, simply because if the system was to crash, I can still access the
data from booting to Dos. I wonder if it would make the most sense to
set the C: drive, to NTFS, keep that a small partition, and use it
mostly just for the OS and programs. Then use Fat32 partitions and/or a
second physicla HD for data storage, and have that set to Fat32???

NTFS isn't available for W98 or ME. IIRC, it came into general use for
the consumer with W2000 and XP. Various versions of it have been around
since the mid-90s, but only for Windows NT which W2000 and XP were
based off of.

Stef

 

[micky]

NTFS isn't available for W98 or ME.

Yeah, if you read his question again, you'll see that that's not what he
meant.

He wants to us NTFS for the operating system XP, in a small partition,
and FAT32 for data which will be accessible both from XP in normal use
and from DOS when things are going badly. I've wondered about that
too, whether that would be a good idea.

One way around it though is to get Hirem's CD or something similar, so
that one can boot to a small version of XP from the CD drive and then
access all your files from that, looking at them on the screen** ,
copying them to a flashdrive, and I'm sure it includes writing files to
a CD or DVD if you have a CD or DVD writer. (you may have to change
boot order in the bios, but you can do that when Windows isn't working)

**(although perhaps not formatted as the program that was meant to read
them would, especially if it's an obscure format, unless that program is
available in a portable version that will run under FE-XP (FE? I
forget the initials for the Hirems version))

A second way around it is to leave the computer off, open it up, unplug
the connectors from the HDD, and use some 25 dollar hardware I just
wrote about to connect the HDD to the USB port of another computer. Of
course that requires two computers.

But I'm still interested in Casey's idea.

 

[micky]

micky formulated the question :

The only person who will know the answer to that is you. Give it a try
and let us know how it turns out.

Well before I read Paul's post that is right below (or above) yours, I
did all this:

Before I read your post, I dl'd Slim Defraggler by Piriform from
<http://www.piriform.com/defraggler/builds> since Rasta posted first,
and I installed it and barely started it up just to look at it. Even
without my doing anything, it filled in the drive representation all in
grey. It had a Benchmark button, and I figured I could do that and
save time later. But when I came back to the program later, it had
done most of the Analysing without my telling it to.

It finished Analysing without stopping at the bad file, whatever that
means!!! I didn't know its full name and I hadn't tried to delete it.
(BTW, Paul, I did look at the name of the file the previous day, and it
was from a file image of my winME computer, which I'll never install
again. Not a data file and I certainly don't need it, and I thought
that was all that mattered, even later when I read your post. But after
I thought about it, I realized it might be a sign of some bigger problem
like you said.)

Here's the link http://www.auslogics.com/en/software/disk-defrag/

Then I downloaded this, mostly because it had a better webpage and
promised more iirc. And had nice pictures of Australia. That's
important. Again I installed it and started it up just to look at it,
and then I thought, well, why not run Analyse again! And it ran in 2
or 3 minutes for 200 GB, even though Defraggler and XP defrag took about
an hour each. Now how can that be?


And I started the defrag, but then I DL'd read Paul's post and stopped
it. Now that Chkdsk runs automatically after certain bad things, and
because with win98 I had a GUI Norton version of CHKDSK, I couldn't
remember what parms to use for Chkdsk but I read XP Help and used /f and
/r. Of course it said I had to restart the computer because it
couldn't lock the drive then. So I restarted, and it ran much slower
than when it decides to run itself. And it checked the free space too,
which took another 45 or so minutes and found NO problems anywhere.
Hooray.

So Auslogic Disk Defrag is running now.

Portable version is near the bottom of the page on the right.

I may get that later. Thanks.

 

[micky]

Um, don't do that!

Thanks for the warning. Details are in my answer to Zo.

Any time data-intensive things are run, even the tools
themselves will use CHKDSK to determine whether it's "safe"
or not. For example, Partition Managers which are about to
resize things, check whether the file system is intact
first, before doing anything dangerous.

The same logic should apply to defragmenters. You shouldn't
run a defragmenter, unless the file system is solid.

Good to know. I'm going to fix up (hopefully) a friend's friend's
computer Sunday, including defrag, based on what I learned tonight.

I would say you have a much more serious problem than
making the partition pretty, and should attend to that error first.
If you think deleting the file is going to fix it, go right
ahead. My guess is, it'll still be broken when you're done.
Run CHKDSK again and see.

While you're at it, you might check your SMART statistics,

Defraggler checks that and it was GOOD.

and see if the hardware is good. It could be, that the
error detected, is related to actual bad hardware. And
you don't want to use either CHKDSK or defragmenters
on flaky hardware. Install a new disk, before letting
those kinds of tools loose.

(The ones in red boxes, should be zero.)

http://oi62.tinypic.com/2wghquh.jpg

Since this has *never* happened to me here, I'd be at least
a bit curious as to the root cause. You might actually
be in serious trouble (a day away from losing a hard drive).

Paul

Thans for the info.

 

[Todd]

Does it

1) condense (defragment free space)?

2) does it come with any junkware (toolbars, pop ups
for crap to buy,etc.)?

Many thanks,
-T

 

[J. P. Gilliver (John)]

In message <[email protected]>, Paul <[email protected]>
writes:
[]

While you're at it, you might check your SMART statistics,
and see if the hardware is good. It could be, that the
error detected, is related to actual bad hardware. And
you don't want to use either CHKDSK or defragmenters
on flaky hardware. Install a new disk, before letting
those kinds of tools loose.

(The ones in red boxes, should be zero.)

http://oi62.tinypic.com/2wghquh.jpg

I see from that that the ones in red boxes are the "Data" column for
"Current Pending Sector", and "Reallocated Sector Cou[nt?]".

CPS is indeed 0 for me, but (using DiskCheckup from www.passmark.com,
but the display is almost identical) RSC isn't. However, one tab ("SMART
Info") I have shows that line as
ID 5
Description Reallocated Sector Count
Status OK
Value 94
Worst 94
Threshold 10
Raw Value 59
Predicted TEC Date 20 Jul 2088

Now, as the values are higher than the threshold, presumably this means
that lower is _worse_, whereas you say they should be zero. I'm puzzled.

Another tab shows (with a drop-down list from which I select the
parameter) what the parameter has been when I've run the utility in the
past. That shows (both "Value" and "Worst", both always being the same)
as 98 from 2010-8-15 (presumably the first time I ever ran it) to
2010-9-19, then (I seem not to have run the utility for a couple of
years) 94 from 2012-3-24, throughout 2013, to still being 94 today. (I'm
_guessing_ I may have done something - maybe dropped the netbook or let
it get very hot? - at some point between 2010-9-19 and 2012-3-24, and
thus damaged a sector or 4?)
[]
So I'm assuming it would never spontaneously _improve_, so 94 is worse
than 98, so I certainly _don't_ want a zero in that column.

What does "Reallocated Sector Count" _mean_ - swapped-out sectors? How
many of the ones the manufacturer originally allocated as spares are
left? (This is what I'd assume it means in my case.)

I suspect different manufacturers - and/or SMART utility programmers -
use these things in different ways, some going down (how many spares are
left) and some going up (how many have been used).

In this case, it _does_ show that one line in red, but I suspect that's
because it's the only one with a Predicted TEC date of other than
"N.A.", presumably because the value is 94 and was once 98 whereas
perhaps none of the other parameters have changed (over the time I've
been running the utility at least)?. I've just re-run the utility, and
the values are the same, other than that the failure (?) date is now
predicted as 2088-7-21, i. e. a day later, presumably based on a
three-point graph using first value, current value, and threshold.

(I'm also puzzled what the difference is between "Value", "Worst", and
"Raw Value"; "Worst" I assume is the worst it's ever been as recorded
inside the drive, but I'm not sure about the other two.)
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

I remember a lot of questions on a vocalist forum about the problems singing
"There is a balm in Gilead" without making it sound like a security alert. -
Linda Fox in UMRA, 2010-11-19

 

[Paul]

While you're at it, you might check your SMART statistics,
and see if the hardware is good. It could be, that the
error detected, is related to actual bad hardware. And
you don't want to use either CHKDSK or defragmenters
on flaky hardware. Install a new disk, before letting
those kinds of tools loose.

(The ones in red boxes, should be zero.)

http://oi62.tinypic.com/2wghquh.jpg

I see from that that the ones in red boxes are the "Data" column for
"Current Pending Sector", and "Reallocated Sector Cou[nt?]".

CPS is indeed 0 for me, but (using DiskCheckup from www.passmark.com,
but the display is almost identical) RSC isn't. However, one tab ("SMART
Info") I have shows that line as
ID 5
Description Reallocated Sector Count
Status OK
Value 94
Worst 94
Threshold 10
Raw Value 59
Predicted TEC Date 20 Jul 2088

Now, as the values are higher than the threshold, presumably this means
that lower is _worse_, whereas you say they should be zero. I'm puzzled.

Another tab shows (with a drop-down list from which I select the
parameter) what the parameter has been when I've run the utility in the
past. That shows (both "Value" and "Worst", both always being the same)
as 98 from 2010-8-15 (presumably the first time I ever ran it) to
2010-9-19, then (I seem not to have run the utility for a couple of
years) 94 from 2012-3-24, throughout 2013, to still being 94 today. (I'm
_guessing_ I may have done something - maybe dropped the netbook or let
it get very hot? - at some point between 2010-9-19 and 2012-3-24, and
thus damaged a sector or 4?)
[]
So I'm assuming it would never spontaneously _improve_, so 94 is worse
than 98, so I certainly _don't_ want a zero in that column.

What does "Reallocated Sector Count" _mean_ - swapped-out sectors? How
many of the ones the manufacturer originally allocated as spares are
left? (This is what I'd assume it means in my case.)

I suspect different manufacturers - and/or SMART utility programmers -
use these things in different ways, some going down (how many spares are
left) and some going up (how many have been used).

In this case, it _does_ show that one line in red, but I suspect that's
because it's the only one with a Predicted TEC date of other than
"N.A.", presumably because the value is 94 and was once 98 whereas
perhaps none of the other parameters have changed (over the time I've
been running the utility at least)?. I've just re-run the utility, and
the values are the same, other than that the failure (?) date is now
predicted as 2088-7-21, i. e. a day later, presumably based on a
three-point graph using first value, current value, and threshold.

(I'm also puzzled what the difference is between "Value", "Worst", and
"Raw Value"; "Worst" I assume is the worst it's ever been as recorded
inside the drive, but I'm not sure about the other two.)[/QUOTE]

All I can tell you, is this. This is what I observed, as a drive
went bad here. The Current and Worst appear to be percentages.
The Data field appears to be "counts" of things. I never got
to the Threshold value, so I don't know what happens when you get
there. Maybe the display in HDTune turns yellow or something.
The Threshold value is set by the hard drive design.

Current Worst Threshold Data Status
Reallocated Sector Count 100 100 36 0 OK
Reallocated Sector Count 100 100 36 57 OK
Reallocated Sector Count 98 98 36 104 OK

A reallocated sector, is a sector no longer in service, which is
replaced by a spare sector on the same track or cylinder perhaps.
It causes a slight slowdown, to make reference to it. Once all the
spares are used up in the vicinity, then no more sparing is
offered (for performance reasons). If the spare sector was on
one end of the disk, and the bad sector on the other end of the
disk, the performance would drop below a megabyte per second.

Note that, Reallocated Sector Count is not an honest count. We don't
know what the units of "Data" are. They might not be sectors.
And to avoid customers checking SMART on a new drive, and
returning the product, maybe the first 100,000 spared out
sectors are not shown in the display. When the 100,001th spared
sector operation occurs, maybe Data is set to 1.

If you look at my results, it would suggest the drive has 5000
spare sectors, which is just silly. Back in the days of much
smaller drives, hard drives were leaving the factory with
100,000 sectors already spared out. The number is likely
larger by now. One reason for even attempting to read or
write the hard drive at the factory, is to develop the
factory defect list, so the drive is "stable" when the
customer goes to use it. Defects should only grow slowly
after that. Factory defects would be microscopic defects
in the surface finish (areas of disk that won't hold data).

Paul