PC almost at standstill

[Terry Pinnell]

That's a good start ... er ... middle?
Did you get the service packs installed and graphics and USB drivers
sorted?

Service packs: No, I'm going to see how I get on with this 'raw' XP for a
while. I've always been apprehensive about SP upgrades following a
disaster years ago. Although my confidence improved following a
much-postponed upgrade to SP3 only a few months ago, I think I'll apply
the 'If it ain't broke...' maxim for now. My shed/workshop computing
requirement is pretty modest and I have no internet connection (yet anyway
- I'd like to add wifi access sometime) so I don't see any major benefits.

USB driver issue. That remains a puzzle. I *thought* this PC, ancient
though it is, did have USB 2.0. I certainly had not been getting those
messages about its absence. Is *that* some sort of SP dependency? Is there
a definitive test I can apply to see what my USB capability is right now?

 

[Paul]

Service packs: No, I'm going to see how I get on with this 'raw' XP for a
while. I've always been apprehensive about SP upgrades following a
disaster years ago. Although my confidence improved following a
much-postponed upgrade to SP3 only a few months ago, I think I'll apply
the 'If it ain't broke...' maxim for now. My shed/workshop computing
requirement is pretty modest and I have no internet connection (yet anyway
- I'd like to add wifi access sometime) so I don't see any major benefits.

USB driver issue. That remains a puzzle. I *thought* this PC, ancient
though it is, did have USB 2.0. I certainly had not been getting those
messages about its absence. Is *that* some sort of SP dependency? Is there
a definitive test I can apply to see what my USB capability is right now?

A7A266-E

http://dlcdnet.asus.com/pub/ASUS/mb/socka/m1647/a7a266-e/a7a266e-100.pdf

PDF page 14. Southbridge is M1535D+.

The company "ALI" became "ULI" and a threat to another chipset maker (NVidia).
NVidia bought the company and crushed it. NVidia continues to list
info on the old ALI chipsets, as in here.

http://www.nvidia.com/page/uli_m1535dplus.html

USB 1.1 Host controller

No USB2 for you.

*******

Now, one advantage of me digging around there, is I found a
driver page for M1535D+.

http://www.nvidia.com/page/uli_drivers.html

"Integrated220"

*******

To get USB2, you can use a PCI to USB2 card. Or, they even make
PCI to USB3 cards (consisting of a PCI to PCI Express bridge,
followed by a PCI Express to USB3 chip). You can retrofit a
better USB solution, and WinXP is probably just enough of
an OS to allow you to fully use it.

Example of a USB3 card.

http://www.newegg.com/Product/Product.aspx?Item=N82E16815201061

The back of the card, shows the bridge chip. And bridging is
a way of getting old PCs, to run newer chip types which are
only available in PCI Express interface.

http://images17.newegg.com/is/image/newegg/15-201-061-Z04?$S640W$

A USB2 card, will be a lot cheaper. Could be had for as little
as 1/5th the price. This is an example of a more conventional
card. Only a single chip. I picked this one, because it had
a NEC chip on board.

http://www.newegg.com/Product/Product.aspx?Item=N82E16815201036

Paul

 

[Terry Pinnell]

A7A266-E

http://dlcdnet.asus.com/pub/ASUS/mb/socka/m1647/a7a266-e/a7a266e-100.pdf

PDF page 14. Southbridge is M1535D+.

The company "ALI" became "ULI" and a threat to another chipset maker (NVidia).
NVidia bought the company and crushed it. NVidia continues to list
info on the old ALI chipsets, as in here.

http://www.nvidia.com/page/uli_m1535dplus.html

USB 1.1 Host controller

No USB2 for you.

OK, thanks. ;-(

I wonder why I'm only now getting the messages about that, in a new XP
install?

*******

Now, one advantage of me digging around there, is I found a
driver page for M1535D+.

http://www.nvidia.com/page/uli_drivers.html

"Integrated220"

Many thanks for taking the time and trouble to research this. I'd have had
no idea that page was relevant. ULi? Integrated220? Would the average end
user be expected to know what that's all about and know how to locate the
correct sw to install? Or is it a page exclusively for IT
techies/Administrators?

Anyway, I've now downloaded it and read in the Reeadme.txt that
"Users don't need to decide what driver should be installed before
executing this integrated driver, i.e. it will automatically decide what
drivers need to be installed in the process of installation," which is
reassuring.

*******

To get USB2, you can use a PCI to USB2 card. Or, they even make
PCI to USB3 cards (consisting of a PCI to PCI Express bridge,
followed by a PCI Express to USB3 chip). You can retrofit a
better USB solution, and WinXP is probably just enough of
an OS to allow you to fully use it.

Example of a USB3 card.

http://www.newegg.com/Product/Product.aspx?Item=N82E16815201061

The back of the card, shows the bridge chip. And bridging is
a way of getting old PCs, to run newer chip types which are
only available in PCI Express interface.

http://images17.newegg.com/is/image/newegg/15-201-061-Z04?$S640W$

A USB2 card, will be a lot cheaper. Could be had for as little
as 1/5th the price. This is an example of a more conventional
card. Only a single chip. I picked this one, because it had
a NEC chip on board.

http://www.newegg.com/Product/Product.aspx?Item=N82E16815201036

Thanks, I'll consider that.

--------------------

In parallel with locating the driver for my Sound Blaster card ( if indeed
this PC does have one - I'm beginning to wonder!) I thought I'd try
switching on the onboard audio in the BIOS, which was disabled. But after
rebooting the H/W Wizard failed to find this:

https://dl.dropboxusercontent.com/u/4019461/MultimediaAudioDriverMissing.jpg


However, this morning I've found what appears to be the original CD setup
disc for SB Live 5.1. I also found 3 other potentially relevant setup CDs.
Looks to me like I should discard 'SB Audio PC164V' but which if any of
the two ASUS CDs seems relevant?

https://dl.dropboxusercontent.com/u/4019461/Old-CDs.jpg

----------

One of the reasons I wanted to avoid a Win XP reinstall was that I
anticipated having to remember dozens of tweaks and customizations. That's
proved to be correct. I've sorted most of the obvious stuff so that my
desktop and general environment is looking reasonably familiar again
(apart from having only about 15 programs installed instead of 550) but
there are several irritations remaining.

One is that 'Shortcut to xxxx' prefix. I KNOW they're shortcuts by the
little arrow BL corner! AFAICT they are removable with the MS TweakUI
Powertoy. I duly ran that but got
https://dl.dropboxusercontent.com/u/4019461/Error-1.jpg
I suspect I may not have executed an old enough version, so will search my
ancient downloads for another.

Another trivial one is removing the Windows 'splash' screen, so changing a
Boot.ini line is now on the todo list also.

On a positive note I do have my macro program )Macro Express Pro) and
large library of macros installed, so the familiar hotkeys are actually
doing something now.

 

[Terry Pinnell]

Thanks, duly downloaded.

And ran. The refresh rate may well have been the issue. Still no
re-occurrence since changing it back to 60 Hz.
https://dl.dropboxusercontent.com/u/4019461/MonitorInfo-1.jpg

And for completeness:
https://dl.dropboxusercontent.com/u/4019461/MonitorInfo-2.jpg

 

[Rob Morley]

In parallel with locating the driver for my Sound Blaster card ( if
indeed this PC does have one - I'm beginning to wonder!) I thought
I'd try switching on the onboard audio in the BIOS, which was
disabled. But after rebooting the H/W Wizard failed to find this:

https://dl.dropboxusercontent.com/u/4019461/MultimediaAudioDriverMissing.jpg


However, this morning I've found what appears to be the original CD
setup disc for SB Live 5.1. I also found 3 other potentially relevant
setup CDs. Looks to me like I should discard 'SB Audio PC164V' but
which if any of the two ASUS CDs seems relevant?

You could run PCISniffer to accurately identify your PCI devices
without pulling them to check the board markings. Assuming yours isn't
an ISA card of course.

 

[Paul]

Thanks, I'll consider that.

--------------------

In parallel with locating the driver for my Sound Blaster card ( if indeed
this PC does have one - I'm beginning to wonder!) I thought I'd try
switching on the onboard audio in the BIOS, which was disabled. But after
rebooting the H/W Wizard failed to find this:

https://dl.dropboxusercontent.com/u/4019461/MultimediaAudioDriverMissing.jpg


However, this morning I've found what appears to be the original CD setup
disc for SB Live 5.1. I also found 3 other potentially relevant setup CDs.
Looks to me like I should discard 'SB Audio PC164V' but which if any of
the two ASUS CDs seems relevant?

https://dl.dropboxusercontent.com/u/4019461/Old-CDs.jpg

----------

One of the reasons I wanted to avoid a Win XP reinstall was that I
anticipated having to remember dozens of tweaks and customizations. That's
proved to be correct. I've sorted most of the obvious stuff so that my
desktop and general environment is looking reasonably familiar again
(apart from having only about 15 programs installed instead of 550) but
there are several irritations remaining.

One is that 'Shortcut to xxxx' prefix. I KNOW they're shortcuts by the
little arrow BL corner! AFAICT they are removable with the MS TweakUI
Powertoy. I duly ran that but got
https://dl.dropboxusercontent.com/u/4019461/Error-1.jpg
I suspect I may not have executed an old enough version, so will search my
ancient downloads for another.

Another trivial one is removing the Windows 'splash' screen, so changing a
Boot.ini line is now on the todo list also.

On a positive note I do have my macro program )Macro Express Pro) and
large library of macros installed, so the familiar hotkeys are actually
doing something now.

This is the answer for TweakUI. It's the Service Pack level of your install.

http://www.theeldergeek.com/forum/index.php?showtopic=6487

"Note that this utility requires you to have upgraded Windows XP to
at least Service Pack 1 (SP1) level. If you try to install it
without SP1, you will get this bizarre error message: "The procedure
entry point GetDllDirectoryW could not be located in the dynamic
link library KERNEL32.dll."
"

You should be able to download the redistributable version of SP3
from Microsoft. Since you're in the UK, I doubt my bookmarks
would be of much good. You can use these as indicators of
how big a download to expect. I checked the link for SP2,
and something started downloading, so you might be able to
find versions of either.

WindowsXP-KB835935-SP2-ENU.exe 278,927,592 bytes
WindowsXP-KB936929-SP3-x86-ENU.exe 331,805,736 bytes

Since you're in the shed, you might not be doing Windows
Update in there, and bringing the redistributable into
the shed on a USB stick, is better than nothing.

Paul

 

[Paul]

And ran. The refresh rate may well have been the issue. Still no
re-occurrence since changing it back to 60 Hz.
https://dl.dropboxusercontent.com/u/4019461/MonitorInfo-1.jpg

And for completeness:
https://dl.dropboxusercontent.com/u/4019461/MonitorInfo-2.jpg

60Hz on an LCD is perfectly adequate, due to the different
persistence properties of LCD pixels. There's no need for
75Hz.

75Hz or 85Hz were more popular in the CRT era. To reduce flicker.

Paul

 

[J. P. Gilliver (John)]

In message <[email protected]>, Terry Pinnell

Thanks a bunch, Paul, appreciate the thorough reply. But I guess you
didn't see my comment about my low skill level in an earlier post:
"Whatever little I ever once knew about this stuff, is now a blurry
memory. Primary and Secondary Channels? Master and Slave? etc, etc? IOW,
I'm no techie ;-)"

So your "Simple" made me smile!

They do me, too!

Paul's replies are excellent and comprehensive (if a bit overwhelming at
times), nicely laid out, with links where appropriate. And I consider
myself a moderate techie, though out of touch with developments in SATA
(and having little knowledge of RAID).

Specifically on master, slave, primary, and secondary: motherboards of
the vintage I think yours is (and which is still the sort I'm most at
ease with!) have (sometimes in addition to some of the newer SATA
interfaces) two EIDE connectors for drives - which are primary and
secondary channels. Each of these can be connected to up to two drives
(i. e. the ribbon cable will have up to three connectors, including the
one that goes into the motherboard); the drives on either channel can be
master and slave, determined by the position (or sometimes
presence/absence) of a link on the drive itself. Blurry memory coming
back? [Sometimes it's done by "cable select", though in my experience
that tends to be more for floppy drives.]

I actually replied some hours ago via my iPad while in the shed but it
seems to have got lost. Anyway things have moved on now.

The current status is that after disconnecting the lower HD, leaving
partitions C and N, I have regained normal performance - and, of course,

Yes, I remember helping with a PC where the drive was failing to the
extent that the PC took 15 minutes to boot, and was intermittently
glacial thereafter. No actual indication of errors though - it just went
_very_ slow! The givaway was that whenever you did anything that
involved the HD (which you could tell from the HD light on the front
coming on), it slowed to a crawl, but anything you did which ran in RAM
was fine. IIRR, emails were mostly fine, once the email prog. itself had
loaded, which confused us a bit!
[]

 

[J. P. Gilliver (John)]

In message <[email protected]>, Terry Pinnell

BTW, after the last crash/reboot I reduced the resolution a bit, to 1024 x
768 and FWIW in the many hours since then haven't had a repetition.

Also, I wonder if monitor refresh rate influences it?

From your use of "a bit", I deduce you reduced it by a factor of less
than one.

Can we assume that - perhaps to keep the shed warm (-:! - it's a CRT
monitor? If it's an LCD (or anything else with discrete pixels, such as
a plasma), I'd generally never run it at other than its native
resolution (or, in extreme cases related to eyesight, a sub-multiple
thereof), as you get quite a bit of blurring.

 

[J. P. Gilliver (John)]

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

You could have at least tried to look at the SMART stats,
to see how bad it is. Look for "Reallocated Sector Count"
and "Current Pending Sectors". A non-zero Data field for
those, is generally a bad sign. HDTune can read out SMART,
via the "Health" tab.

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

Paul

"Reallocated Sector Count" (and possibly the other one) seems to be
interpreted different ways by different utilities, and/or used
differently by different manufacturers. I don't have hdtune (or don't
think I have), but I have two other utilities.

Seagate's SeaTools for Windows, under Basic Tests, includes "S.M.A.R.T
Check", which tells me "Congratulations, your Seagate product has passed
an important diagnostic test", and its log file shows that I've run
SMART a few times since 2012, as well as other tests it offers - Short
DST, Short Generic, and Long Generic (which looks to have taken just
under an hour - and not get any longer, 56'29" both times - both times I
ran it, 2012-2-25 and 2012-5-2). It doesn't give any details of the
SMART test other than Pass - fair enough, considering it isn't a Seagate
drive! (Unless they include Samsung, which it is.)

DiscCheckupV3.0 (http://www.passmark.com/products/diskcheckup.htm, now
at 3.1 I see) lists assorted parameters, with status, value, worst,
threshold, raw value, and predicted TEC Date. This last is shown as N.A.
for all parameters except Reallocated Sector Count, presumably because
the others all (other than Temperature) have the same value under Value
(which I take to mean what it's just been read as) as under Worst.

For "5 Reallocated Sector Count", it shows Status: OK (as it does for
everything else), Value: 94, Worst: 94, Threshold: 10, Raw Value: 59,
Predicted TEC Date: 10 Jul 2072 20:44:11. For "C5 Current Pending
Sector Count", it shows OK, 252, 252, 0, 0, and N.A.

It also has a "SMART History" tag, where I can examine what all the
parameters were for all the times I've run it in the past (it says
Records Found 22, presumably meaning I've run it that many times
including just now). For RS Count, it shows a "Value" and "Worst" of
(both) 98 on 16 occasions from 2010-7-15 to 2010-9-19, and (both) 94 on
6 occasions from 2012-3-24 to today; for CPS Count, it shows Value and
Worst of 252 on all 22 occasions.

From this, I deduce that RS Count is something that this utility is
expecting to go _down_ as things deteriorate, with the threshold of 0
supporting this. Actually, clicking on Help takes me to an excellent
(HTML-based; it's on my disc, otherwise I'd give a link) help section,
which tells me that item 5 "Represents the amount of spare sector pool
available."

 

[Paul]

You could have at least tried to look at the SMART stats,
to see how bad it is. Look for "Reallocated Sector Count"
and "Current Pending Sectors". A non-zero Data field for
those, is generally a bad sign. HDTune can read out SMART,
via the "Health" tab.

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

Paul

"Reallocated Sector Count" (and possibly the other one) seems to be
interpreted different ways by different utilities, and/or used
differently by different manufacturers. I don't have hdtune (or don't
think I have), but I have two other utilities.

Seagate's SeaTools for Windows, under Basic Tests, includes "S.M.A.R.T
Check", which tells me "Congratulations, your Seagate product has passed
an important diagnostic test", and its log file shows that I've run
SMART a few times since 2012, as well as other tests it offers - Short
DST, Short Generic, and Long Generic (which looks to have taken just
under an hour - and not get any longer, 56'29" both times - both times I
ran it, 2012-2-25 and 2012-5-2). It doesn't give any details of the
SMART test other than Pass - fair enough, considering it isn't a Seagate
drive! (Unless they include Samsung, which it is.)

DiscCheckupV3.0 (http://www.passmark.com/products/diskcheckup.htm, now
at 3.1 I see) lists assorted parameters, with status, value, worst,
threshold, raw value, and predicted TEC Date. This last is shown as N.A.
for all parameters except Reallocated Sector Count, presumably because
the others all (other than Temperature) have the same value under Value
(which I take to mean what it's just been read as) as under Worst.

For "5 Reallocated Sector Count", it shows Status: OK (as it does for
everything else), Value: 94, Worst: 94, Threshold: 10, Raw Value: 59,
Predicted TEC Date: 10 Jul 2072 20:44:11. For "C5 Current Pending
Sector Count", it shows OK, 252, 252, 0, 0, and N.A.

It also has a "SMART History" tag, where I can examine what all the
parameters were for all the times I've run it in the past (it says
Records Found 22, presumably meaning I've run it that many times
including just now). For RS Count, it shows a "Value" and "Worst" of
(both) 98 on 16 occasions from 2010-7-15 to 2010-9-19, and (both) 94 on
6 occasions from 2012-3-24 to today; for CPS Count, it shows Value and
Worst of 252 on all 22 occasions.

From this, I deduce that RS Count is something that this utility is
expecting to go _down_ as things deteriorate, with the threshold of 0
supporting this. Actually, clicking on Help takes me to an excellent
(HTML-based; it's on my disc, otherwise I'd give a link) help section,
which tells me that item 5 "Represents the amount of spare sector pool
available."[/QUOTE]

This is a sequence of values from a disk in declining health.
As recorded in HDTune.

It implies the "total available health" is around (100/100-98)*104 = 5000
Of which at this point, the disk has used up 104 of the 5000 maximum
allowed.
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

And that is not a "honest" indicator. The disk manufacturer does
not want customers receiving a new disk, and seeing a non-zero "Data"
value. So, the value is actually thresholded in some way. The disk
has a growing error count, but nothing shows in Data. Then, one day,
you start to see counts there, meaning you're just past the
threshold. Health seems to go downhill fast. And it could be just
the nature of the thresholding that makes it seem that way.

In any case, none of the drives I've caught, showing a count,
have died on me yet. So swapping out the drive when it
starts to indicate a problem, seems to head off a disaster
in time. Your mileage may vary of course.

My experience with the Maxtor was, death occurred soon after
symptoms (same day). But back in that era, I wasn't using SMART,
so I have no idea if "Reallocated" had been giving any warning or
not.

According to official study, SMART isn't that good at catching
stuff in time. So don't rely on studious usage of SMART
as a substitute for doing backups! I.e. Don't wait until
reallocated shows 98% health, before you do your first
backup. You should be doing the backups all along, as disks
do die without any prior warning in SMART. Seeing as some
failures are traceable to data structure corruption, not
all the failures are media (dirty platter) related.

Paul

 

[Terry Pinnell]

You could run PCISniffer to accurately identify your PCI devices
without pulling them to check the board markings. Assuming yours isn't
an ISA card of course.

Thanks, I'll try that.

 

[Terry Pinnell]

This is the answer for TweakUI. It's the Service Pack level of your install.

http://www.theeldergeek.com/forum/index.php?showtopic=6487

"Note that this utility requires you to have upgraded Windows XP to
at least Service Pack 1 (SP1) level. If you try to install it
without SP1, you will get this bizarre error message: "The procedure
entry point GetDllDirectoryW could not be located in the dynamic
link library KERNEL32.dll."
"

You should be able to download the redistributable version of SP3
from Microsoft. Since you're in the UK, I doubt my bookmarks
would be of much good. You can use these as indicators of
how big a download to expect. I checked the link for SP2,
and something started downloading, so you might be able to
find versions of either.

WindowsXP-KB835935-SP2-ENU.exe 278,927,592 bytes
WindowsXP-KB936929-SP3-x86-ENU.exe 331,805,736 bytes

Since you're in the shed, you might not be doing Windows
Update in there, and bringing the redistributable into
the shed on a USB stick, is better than nothing.

Paul

Thanks, well found!

 

[J. P. Gilliver (John)]

You could have at least tried to look at the SMART stats, []
http://www.hdtune.com/files/hdtune_255.exe

[/QUOTE][/QUOTE]

I've now downloaded (and run) it: nice clean presentation.

Paul

"Reallocated Sector Count" (and possibly the other one) seems to be
interpreted different ways by different utilities, and/or used []
DiscCheckupV3.0 (http://www.passmark.com/products/diskcheckup.htm, []
For "5 Reallocated Sector Count", it shows Status: OK (as it does
for everything else), Value: 94, Worst: 94, Threshold: 10, Raw Value:
59, Predicted TEC Date: 10 Jul 2072 20:44:11. For "C5 Current
Pending Sector Count", it shows OK, 252, 252, 0, 0, and N.A.

[]
This is a sequence of values from a disk in declining health.

I was worried there, thinking you meant my drive ...

As recorded in HDTune.

.... until I read that! You mean that what follows is an example.

It implies the "total available health" is around (100/100-98)*104 = 5000
Of which at this point, the disk has used up 104 of the 5000 maximum
allowed.
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

(How did you know it was 5000?)

And that is not a "honest" indicator. The disk manufacturer does
not want customers receiving a new disk, and seeing a non-zero "Data"
value. So, the value is actually thresholded in some way. The disk
has a growing error count, but nothing shows in Data. Then, one day,
you start to see counts there, meaning you're just past the
threshold. Health seems to go downhill fast. And it could be just
the nature of the thresholding that makes it seem that way.

Hmm. Diskcheckup's "Raw" seems to be the same as HDtune's "Data". HDtune
doesn't (AFAICT) have a past readings log; DiskCheckup only seems to log
the "Value" and "Worst" columns, not the "Raw/Data" one. As I said, it
shows me for RS Count 98 on 16 occasions from 2010-7-15 to 2010-9-19
then 94 on (now 9) occasions from 2012-3-24 to now; I don't remember
what prompted me to run it any time (probably discussions like these),
so I don't know why I didn't run it at all during 2011, nor what
happened to the PC during that time. I'll keep an eye on the 59 value.

In any case, none of the drives I've caught, showing a count,
have died on me yet. So swapping out the drive when it
starts to indicate a problem, seems to head off a disaster
in time. Your mileage may vary of course.
Indeed.

My experience with the Maxtor was, death occurred soon after
symptoms (same day). But back in that era, I wasn't using SMART,
so I have no idea if "Reallocated" had been giving any warning or
not.

According to official study, SMART isn't that good at catching
stuff in time. So don't rely on studious usage of SMART
as a substitute for doing backups! I.e. Don't wait until

Indeed not! I have my most irreplaceable data copied onto my brother's
machine a long way away, and other data gets copied to CD from time to
time.

reallocated shows 98% health, before you do your first

By "Health", do you mean the "Current" and "Worst" values? The only
mention of "Health" in HD Tune that I can see is "Health Status: Ok" in
green at the bottom right, which is of course not a numerical value. But
I don't think "Current" and "Worst" can be percentage values, since for
several parameters they are 252, and for one (E1 Load/Unload Cycle
Count) they're 2, and it shows Ok under status for all of them.

backup. You should be doing the backups all along, as disks
do die without any prior warning in SMART. Seeing as some

I know - I've had one go like that. Though I think it was a couple of
decades ago! (IME, solid state memory - such as USB sticks - is far more
prone to die suddenly and completely.)

failures are traceable to data structure corruption, not
all the failures are media (dirty platter) related.

Indeed. (Though believing those - data structure corruptions - can be
recovered from is dangerous too! For many of us it's still lost.)

Paul

--?

 

[Rob Morley]

BTW, after the last crash/reboot I reduced the resolution a bit, to
1024 x 768 and FWIW in the many hours since then haven't had a
repetition.

Also, I wonder if monitor refresh rate influences it?

Refresh rate can make a difference, 60Hz is fine on an LCD monitor but
you really don't want to be using anything other than native resolution
if the graphics card will handle it.

 

[Bob F]

Terry Pinnell wrote:
----------

One of the reasons I wanted to avoid a Win XP reinstall was that I
anticipated having to remember dozens of tweaks and customizations.
That's proved to be correct. I've sorted most of the obvious stuff so
that my desktop and general environment is looking reasonably
familiar again (apart from having only about 15 programs installed
instead of 550) but there are several irritations remaining.

One is that 'Shortcut to xxxx' prefix. I KNOW they're shortcuts by the
little arrow BL corner! AFAICT they are removable with the MS TweakUI
Powertoy. I duly ran that but got
https://dl.dropboxusercontent.com/u/4019461/Error-1.jpg
I suspect I may not have executed an old enough version, so will
search my ancient downloads for another.

You can just right click on the shortcut and rename it.

Another trivial one is removing the Windows 'splash' screen, so
changing a Boot.ini line is now on the todo list also.

On a positive note I do have my macro program )Macro Express Pro) and
large library of macros installed, so the familiar hotkeys are
actually doing something now.

All this is the main resaon I frequently make image backups of my C drive. Ican
easily restore my system to a recent known good state if the drive dies of an
infestation trashes it. It takes maybe 1/2 hour to restore my OS using the
image.

 

[Paul]

In message <[email protected]>, Paul <[email protected]>
writes:
[]
You could have at least tried to look at the SMART stats, []
http://www.hdtune.com/files/hdtune_255.exe

I've now downloaded (and run) it: nice clean presentation.

Paul
"Reallocated Sector Count" (and possibly the other one) seems to be
interpreted different ways by different utilities, and/or used []
DiscCheckupV3.0 (http://www.passmark.com/products/diskcheckup.htm, []
For "5 Reallocated Sector Count", it shows Status: OK (as it does
for everything else), Value: 94, Worst: 94, Threshold: 10, Raw
Value: 59, Predicted TEC Date: 10 Jul 2072 20:44:11. For "C5
Current Pending Sector Count", it shows OK, 252, 252, 0, 0, and N.A.

[]
This is a sequence of values from a disk in declining health.

I was worried there, thinking you meant my drive ...

As recorded in HDTune.

... until I read that! You mean that what follows is an example.

It implies the "total available health" is around (100/100-98)*104 = 5000
Of which at this point, the disk has used up 104 of the 5000 maximum
allowed.
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

(How did you know it was 5000?)

And that is not a "honest" indicator. The disk manufacturer does
not want customers receiving a new disk, and seeing a non-zero "Data"
value. So, the value is actually thresholded in some way. The disk
has a growing error count, but nothing shows in Data. Then, one day,
you start to see counts there, meaning you're just past the
threshold. Health seems to go downhill fast. And it could be just
the nature of the thresholding that makes it seem that way.

Hmm. Diskcheckup's "Raw" seems to be the same as HDtune's "Data". HDtune
doesn't (AFAICT) have a past readings log; DiskCheckup only seems to log
the "Value" and "Worst" columns, not the "Raw/Data" one. As I said, it
shows me for RS Count 98 on 16 occasions from 2010-7-15 to 2010-9-19
then 94 on (now 9) occasions from 2012-3-24 to now; I don't remember
what prompted me to run it any time (probably discussions like these),
so I don't know why I didn't run it at all during 2011, nor what
happened to the PC during that time. I'll keep an eye on the 59 value.

In any case, none of the drives I've caught, showing a count,
have died on me yet. So swapping out the drive when it
starts to indicate a problem, seems to head off a disaster
in time. Your mileage may vary of course.
Indeed.

My experience with the Maxtor was, death occurred soon after
symptoms (same day). But back in that era, I wasn't using SMART,
so I have no idea if "Reallocated" had been giving any warning or
not.

According to official study, SMART isn't that good at catching
stuff in time. So don't rely on studious usage of SMART
as a substitute for doing backups! I.e. Don't wait until

Indeed not! I have my most irreplaceable data copied onto my brother's
machine a long way away, and other data gets copied to CD from time to
time.

reallocated shows 98% health, before you do your first

By "Health", do you mean the "Current" and "Worst" values? The only
mention of "Health" in HD Tune that I can see is "Health Status: Ok" in
green at the bottom right, which is of course not a numerical value. But
I don't think "Current" and "Worst" can be percentage values, since for
several parameters they are 252, and for one (E1 Load/Unload Cycle
Count) they're 2, and it shows Ok under status for all of them.

backup. You should be doing the backups all along, as disks
do die without any prior warning in SMART. Seeing as some

I know - I've had one go like that. Though I think it was a couple of
decades ago! (IME, solid state memory - such as USB sticks - is far more
prone to die suddenly and completely.)

failures are traceable to data structure corruption, not
all the failures are media (dirty platter) related.

Indeed. (Though believing those - data structure corruptions - can be
recovered from is dangerous too! For many of us it's still lost.)

I figure these things out, by looking at the trends. "Current" is in
units of percentage. When it reads 98 percent, I've lost 2 percent
of the available reallocation sectors. And that was equal to somewhere
around 104 sectors. If we do the math, the available pool of spares
must be around 5000. (Of course, this is all sheer imagination,
as we don't really know what it is counting, and what it is
indicating. Don't expect documentation or anything.)

As I indicated in my answer, when the drive was new, the statistics
would be "spotless", so customers would not be returning the drives.
And that means, the indicators can't be completely linear.
A little "defect hiding" goes on, to make the disks look
spotless when shipped. I can assure you, that no disk is
free of reallocations, when it ships. (On SCSI drives, you
can pull the factory defect list from the drive, which is what I used
to do for drives at work. None of the drives shipped with
no reallocations. There are always some. I used to print off
the factory defect list, and tape the sheets of paper inside
the computer for later. SCSI drives allow resetting reallocations
if you want, and allowing the drive to rediscover them again.)

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

In HDTune, the Data column will have increasing values. The
percentage-based Current, will have a declining value. A bad spot
could develop in the disk, well before the "life" of the drive
is exhausted, so that metric above may not have any relationship
to reality.

Since reallocations cannot be reset on IDE drives, if the raw count
is 104 today, it should not be 103 tomorrow. It is supposed
to only increase. Sector sparing on IDE is supposed to be
totally automated, and out of the user's control. Only
accessing the secret three pin TTL level serial interface
on the drive controller, and issuing some undocumented command,
could change that.

Paul

 

[Jon Danniken]

Since reallocations cannot be reset on IDE drives, if the raw count
is 104 today, it should not be 103 tomorrow. It is supposed
to only increase. Sector sparing on IDE is supposed to be
totally automated, and out of the user's control. Only
accessing the secret three pin TTL level serial interface
on the drive controller, and issuing some undocumented command,
could change that.

It has always bugged me to see unused pin headers inside of different
pieces of hardware and not be able to look at documentation for what can
be done with them. Bastards.

Jon

 

[Paul]

It has always bugged me to see unused pin headers inside of different
pieces of hardware and not be able to look at documentation for what can
be done with them. Bastards.

Jon

Looks like this article, is connecting to pins on the back.

http://www.msfn.org/board/topic/128807-the-solution-for-seagate-720011-hdds/

And I don't think I've seen any document that addresses the commands
they were using there. "m0,2,2,0,0,0,0,22 <enter>"

Perhaps I was confusing that procedure, with the one for some router, where
you connect a serial port to some pads on the surface of the PCB itself.

Paul

 

[Jon Danniken]

Looks like this article, is connecting to pins on the back.

http://www.msfn.org/board/topic/128807-the-solution-for-seagate-720011-hdds/


And I don't think I've seen any document that addresses the commands
they were using there. "m0,2,2,0,0,0,0,22 <enter>"

Perhaps I was confusing that procedure, with the one for some router, where
you connect a serial port to some pads on the surface of the PCB itself.

My guess is someone with access (like a seagate tech) volunteered that
information to help techs with a specific issue. It sure would be fun
to know what all is available, though, if only to play around with stuff.

I have seen people solder on pin headers to routers in order to
repurpose them, but in those cases it was pretty clear it was just a
regular RS232 port on Linux firmware.

Jon