HD write speed slows to a crawl - why?

Running XP, SP3, all updates current. This started all of a sudden, and
two different HD diagnostics (one from the manufacturer) show no
problem. Benchmark test shows read speed from HD to be normal, but
*write* speed is now about .54 MB per second (a crawl). All virus and
malware scans show no problems at all. Any advice much appreciated!

Tony

In ,
Tony typed:
Running XP, SP3, all updates current. This started all
of a sudden, and two different HD diagnostics (one from
the manufacturer) show no problem. Benchmark test shows
read speed from HD to be normal, but *write* speed is now
about .54 MB per second (a crawl). All virus and malware
scans show no problems at all. Any advice much
appreciated!
Tony

I hope someone else has better advice for you, but I can surmise this:
The problem isn't hard drive related per your description, so at least it
seems to have been eliminated as the root cause.

Actually, more details would help a LOT.

How much RAM?

Boot drive size?

Run Defrag? That likely won't fix anything but it can have an impact on data
thru-put.

What antivirus and antimalware programs have you run?
Did you update them just before running them so you have the latest files
etc. for them? The files change often and it's important to know what you
did run.

Have you run, from a Command Prompt (DOS window to some)
chkdsk (boot drive): /f ?
If not, you should do so.

Have you tried using a System Restore Point with a date earlier than when
the problem first appeared?

Have you installed anything recently?
Have you uninstalled anything recently?

Is there any unusual disk-thrashing during the slowdowns?

How are you checking the disk for data speeds? What app are you using?

Do you back up regularly?
If so, have you tried to revert your drive to an earlier time?

Have you used MSConfig (Start; Run; type in msconfig and Return)?
It's only for testing, not making permanent changes, but if turning off
everything but the essentials fixes the problem, you have a place to start
working from. If killing them fixes the problem, start turning them back on
one at a time until you find the one causing a problem and report it here.

Details, details, details are the three things that will let others help you
with the least guesswork.

Please report back with the answers to those questions. Better yet, start a
new thread repeating your problem and including the answers to these
questions.

HTH,

Twayne`

On 03/25/2013 07:45 PM, Tony wrote:
Running XP, SP3, all updates current. This started all of a sudden, and
two different HD diagnostics (one from the manufacturer) show no
problem. Benchmark test shows read speed from HD to be normal, but
*write* speed is now about .54 MB per second (a crawl). All virus and
malware scans show no problems at all. Any advice much appreciated!

Tony



I've seen "bad" hard drives pass the diagnostic test.

I'd get a new drive and clone the old one to it

--
https://www.createspace.com/3707686

philo wrote:
On 03/25/2013 07:45 PM, Tony wrote:
Running XP, SP3, all updates current. This started all of a sudden, and
two different HD diagnostics (one from the manufacturer) show no
problem. Benchmark test shows read speed from HD to be normal, but
*write* speed is now about .54 MB per second (a crawl). All virus and
malware scans show no problems at all. Any advice much appreciated!

Tony



I've seen "bad" hard drives pass the diagnostic test.

I'd get a new drive and clone the old one to it


There is a free version of HDTune, suitable for reading out
the SMART statistics.

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

Check your "reallocated sector count" and "current pending sector"
in the SMART statistics. Look under the Data column for each.
The value should be zero if the drive is OK. Those are the two
statistics I rely on here, for considering replacement (i.e.
arbitrary decision point, trading cost of new drive, versus
cost of losing data). There's really no way to predict accurately
when a drive will tip over - I've had a drive die in mid-session,
so they don't even die reliably at startup (normally, you might
discover the drive is bad, first thing in the morning - no boot).

http://forums.vr-zone.com/photopost/...ST3200822A.jpg

This is an example, of three readings of a hard drive here, over
a period of a few days. Notice that the "Current" field, keeps
track of what percentage of life, the "104" in the data column
represents. That means the drive considers about 5000 errors
to be "toast time". If 5000 showed in the Data column, the
percentage of life in Current should be zero.

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

If the drive had a high value for Current Pending (sectors needing
attention), and those sectors were being evaluated on a write, that
might slow writing down.

But as theories go, it's pretty hard to justify *consistent*
differences in behavior between read and write. If everything
was "smooth" and writes were just slow continually, it would be
pretty hard to explain. It's easier for performance to be
"spotty", bad one moment, good the next.

(Note - to see this JPG, you may need to view this page *twice*,
as Photobucket converts a direct link to their JPG, into an HTML
page where you can't see anything. A second click may make the
picture visible.) In this example, the HDTune read benchmark seems to be
showing a "bad spot" near the 70% mark. The spikes near the left
of the graph, could be irregularities in the benchmarking effort,
and it's only when the spikes get really wide, you suspect a bad spot.
You would need to be writing to the bad spot, over and over again,
to get a consistent slowdown.

http://i42.photobucket.com/albums/e3...005/HDTUNE.jpg

*******

A drive interface can slip into PIO mode, which would be
a reason for a "flat" benchmark curve in HDTune. But read
and write should be affected the same way in that case.

http://support.microsoft.com/default...b;en-us;817472

Paul

In message , philo*
writes:
On 03/25/2013 07:45 PM, Tony wrote:
Running XP, SP3, all updates current. This started all of a sudden, and
two different HD diagnostics (one from the manufacturer) show no
problem. Benchmark test shows read speed from HD to be normal, but
*write* speed is now about .54 MB per second (a crawl). All virus and
malware scans show no problems at all. Any advice much appreciated!

Tony



I've seen "bad" hard drives pass the diagnostic test.

I'd get a new drive and clone the old one to it

Seconded. I knew someone whose HD made absolutely no errors, but his
system slowed to a crawl (it was taking a quarter of an hour to boot!)
whenever it accessed the disc; it ran at normal speed when doing
something in memory. Cloning to a new disc (well, some sort of copying;
I used a utility from the [new] manufacturer. The new one was bigger)
restored the computer to normal.
--
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Britain is still a class-ridden society. As soon as a man opens his mouth, we
can tell in what sort of school he missed his education. (George Mikes, "How to
be Decadent" [1977].)

In ,
Tony typed:
Running XP, SP3, all updates current. This started all
of a sudden, and two different HD diagnostics (one from
the manufacturer) show no problem. Benchmark test shows
read speed from HD to be normal, but *write* speed is now
about .54 MB per second (a crawl). All virus and malware
scans show no problems at all. Any advice much
appreciated!
Tony

So far the advice hasn't been much of any help other than to urge you to
spend money, which isn't necessary yet.

If I don't see answers to any of my questions or evidence that you've tried
to turn you computer back via System Restore, a backup or whatever else you
may have, including a full reinstall from the original OD disc, it'll be
pretty obvious you're not actually looking for real help.

Or are you just trolling? Or out of town? What?

HTH,

Twayne`

In message , Twayne
writes:
[]
Have you used MSConfig (Start; Run; type in msconfig and Return)?
It's only for testing, not making permanent changes, but if turning off
everything but the essentials fixes the problem, you have a place to start
working from. If killing them fixes the problem, start turning them back on
one at a time until you find the one causing a problem and report it here.
[]
Rather than turning them on one at a time, a binary search is quicker:
turn half of them back on, and then if the problem does/doesn't recur,
turn off/on half of the ones you turned on/didn't turn on the first
time, ...
--
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

"The people here are more educated and intelligent. Even stupid people in
Britain are smarter than Americans." Madonna, in RT 30 June-6July 2001 (page
32)

On Tuesday, March 26, 2013 5:40:30 AM UTC-5, Paul wrote:
philo wrote:

On 03/25/2013 07:45 PM, Tony wrote:

Running XP, SP3, all updates current. This started all of a sudden, and

two different HD diagnostics (one from the manufacturer) show no

problem. Benchmark test shows read speed from HD to be normal, but

*write* speed is now about .54 MB per second (a crawl). All virus and

malware scans show no problems at all. Any advice much appreciated!



Tony







I've seen "bad" hard drives pass the diagnostic test.



I'd get a new drive and clone the old one to it





There is a free version of HDTune, suitable for reading out

the SMART statistics.



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



Check your "reallocated sector count" and "current pending sector"

in the SMART statistics. Look under the Data column for each.

The value should be zero if the drive is OK. Those are the two

statistics I rely on here, for considering replacement (i.e.

arbitrary decision point, trading cost of new drive, versus

cost of losing data). There's really no way to predict accurately

when a drive will tip over - I've had a drive die in mid-session,

so they don't even die reliably at startup (normally, you might

discover the drive is bad, first thing in the morning - no boot).



http://forums.vr-zone.com/photopost/...ST3200822A.jpg



This is an example, of three readings of a hard drive here, over

a period of a few days. Notice that the "Current" field, keeps

track of what percentage of life, the "104" in the data column

represents. That means the drive considers about 5000 errors

to be "toast time". If 5000 showed in the Data column, the

percentage of life in Current should be zero.



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



If the drive had a high value for Current Pending (sectors needing

attention), and those sectors were being evaluated on a write, that

might slow writing down.



But as theories go, it's pretty hard to justify *consistent*

differences in behavior between read and write. If everything

was "smooth" and writes were just slow continually, it would be

pretty hard to explain. It's easier for performance to be

"spotty", bad one moment, good the next.



(Note - to see this JPG, you may need to view this page *twice*,

as Photobucket converts a direct link to their JPG, into an HTML

page where you can't see anything. A second click may make the

picture visible.) In this example, the HDTune read benchmark seems to be

showing a "bad spot" near the 70% mark. The spikes near the left

of the graph, could be irregularities in the benchmarking effort,

and it's only when the spikes get really wide, you suspect a bad spot.

You would need to be writing to the bad spot, over and over again,

to get a consistent slowdown.



http://i42.photobucket.com/albums/e3...005/HDTUNE.jpg



*******



A drive interface can slip into PIO mode, which would be

a reason for a "flat" benchmark curve in HDTune. But read

and write should be affected the same way in that case.



http://support.microsoft.com/default...b;en-us;817472



Paul

After testing HDDTune and Speedfan, I am wondering how reliable the S.M.A.R.T. info is. I tried to find source code to write a program, but I only found one very old source code.

Lowest reading given by HDDTune was about a month ago and today.
Speedfan gave the higher number.

If that number was valid, I don't think my computer would even boot up.

Andy

(05) Reallocated Sector Count 79 79 24 32901304 Ok
3/27/13 " 5DA01F6088B

Andy wrote:
On Tuesday, March 26, 2013 5:40:30 AM UTC-5, Paul wrote:
philo wrote:

On 03/25/2013 07:45 PM, Tony wrote:
Running XP, SP3, all updates current. This started all of a sudden, and
two different HD diagnostics (one from the manufacturer) show no
problem. Benchmark test shows read speed from HD to be normal, but
*write* speed is now about .54 MB per second (a crawl). All virus and
malware scans show no problems at all. Any advice much appreciated!
Tony
I've seen "bad" hard drives pass the diagnostic test.
I'd get a new drive and clone the old one to it


There is a free version of HDTune, suitable for reading out

the SMART statistics.



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



Check your "reallocated sector count" and "current pending sector"

in the SMART statistics. Look under the Data column for each.

The value should be zero if the drive is OK. Those are the two

statistics I rely on here, for considering replacement (i.e.

arbitrary decision point, trading cost of new drive, versus

cost of losing data). There's really no way to predict accurately

when a drive will tip over - I've had a drive die in mid-session,

so they don't even die reliably at startup (normally, you might

discover the drive is bad, first thing in the morning - no boot).



http://forums.vr-zone.com/photopost/...ST3200822A.jpg



This is an example, of three readings of a hard drive here, over

a period of a few days. Notice that the "Current" field, keeps

track of what percentage of life, the "104" in the data column

represents. That means the drive considers about 5000 errors

to be "toast time". If 5000 showed in the Data column, the

percentage of life in Current should be zero.



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



If the drive had a high value for Current Pending (sectors needing

attention), and those sectors were being evaluated on a write, that

might slow writing down.



But as theories go, it's pretty hard to justify *consistent*

differences in behavior between read and write. If everything

was "smooth" and writes were just slow continually, it would be

pretty hard to explain. It's easier for performance to be

"spotty", bad one moment, good the next.



(Note - to see this JPG, you may need to view this page *twice*,

as Photobucket converts a direct link to their JPG, into an HTML

page where you can't see anything. A second click may make the

picture visible.) In this example, the HDTune read benchmark seems to be

showing a "bad spot" near the 70% mark. The spikes near the left

of the graph, could be irregularities in the benchmarking effort,

and it's only when the spikes get really wide, you suspect a bad spot.

You would need to be writing to the bad spot, over and over again,

to get a consistent slowdown.



http://i42.photobucket.com/albums/e3...005/HDTUNE.jpg



*******



A drive interface can slip into PIO mode, which would be

a reason for a "flat" benchmark curve in HDTune. But read

and write should be affected the same way in that case.



http://support.microsoft.com/default...b;en-us;817472



Paul

After testing HDDTune and Speedfan, I am wondering how reliable the S.M.A.R.T. info is. I tried to find source code to write a program, but I only found one very old source code.

Lowest reading given by HDDTune was about a month ago and today.
Speedfan gave the higher number.

If that number was valid, I don't think my computer would even boot up.

Andy

(05) Reallocated Sector Count 79 79 24 32901304 Ok
3/27/13 " 5DA01F6088B

Is the drive an SSD ?

The definition of the SMART fields changes for SSD, because
they aren't based on platters. The industry chose a new set
of definitions.

Which is one reason why the free version of HDTune isn't suited
for all situations. Since the free version hasn't been updated
for some time, it's going to get out of date.

And I'm not even sure I can track down where the SMART parameters
are documented. While it might be in ATA/ATAPI spec, it could
just as easily have been placed in some kind of separate spec.

http://en.wikipedia.org/wiki/S.M.A.R.T

"As a result, S.M.A.R.T. is not always implemented correctly
on many computer platforms, due to the absence of industry-wide
software and hardware standards for S.M.A.R.T. data interchange."

Based on the number of "yellow" irrelevant marks in HDTune,
I'd have to say that was the case.

I also would not have had a problem with this approach, if say,
WD issued a WD_SMART utility, Seagate a Seagate_SMART utility,
which would correctly identify all the information presented.
Relying on "Joe's Garage" to write a SMART decoder for Seagate,
is just stupid. The poor software developers aren't likely
getting any secret info from the disk companies. The
disk companies should write their own decoders, since they're
making this crap up.

Paul