Failing HDDs

"Cows are Nice" wrote in message
...
Fact - Most dead HDDs, nearly all of them, are found to be infected with
NTFS.

And what proportion of all discs, dead or alive, use NTFS? Could it be that
dead NTFS discs are more common than those with any other filesystem, simply
because NTFS is the most widely-used filesystem - at least on Windows. If
you said that the error rate for NTFS (number of dead as a proportion of
number used) is greater than for any other filesystem, then I'd be more
worried.

Anyway, what do you mean by "dead"? Do you mean at the hardware level
(failed controller board, failed head tracking, failed motor etc) or do you
mean at the logical level (corrupt file-to-storage-sectors or file-to-folder
mapping)? I can see how logical errors could be more prevalent for some
filesystems than others, but I can't see that the choice of filesystem would
affect hardware errors.


What is regarded as the best filesystem (in terms of long
filename/directory/pathnaem support, large file support, reliability,
ability to repair logical errors) that is supported by a wide variety of
operating systems in case you want to use the same portable drive in
Windows, Mac, Linux?

FAT/FAT32 has the 4 GB file size limit, at least as implemented by Windows,
even though not enshrined in the standard. But it is free, which is why it
is still used on cameras etc, even video cameras where it's necessary to
split long recordings into 4 GB chunks.

exFAT isn't universally understood - I *think* it's some flavours of Linux
that don't recognise it - but it does support files 4 GB. Is its
error-recovery and journalling as good as NTFS?

NTFS *appears* to be good, but it has the problem that it's proprietary to
Microsoft. I wonder how Ubuntu and Raspian get round having to pass on a
Microsoft licence fee to users. Reverse-engineered by examining the disc and
working out how it ended up that way?

On 4/26/2019 2:24 PM, Cows are Nice wrote:
On 04/26/2019 10:06 AM, Filip454 wrote:
W dniu 2019-04-26 o 06:31, Cows are Nice pisze:
Windows ruins hardware

Could you elaborate?

Yes. Every so often, an anti-Linux troll gets swatted for visiting Linux
groups to spread alarm that Linux will destroy their hardware. Since we
enjoy that, following up with many witty replies, I thought it a swell
idea to bring the fun to win-droids with a crosspost.

Fact - Most dead HDDs, nearly all of them, are found to be infected with
NTFS.
Linux is a minor player in the personal computer world, so statistically
that is correct. The largest percentage of computers in use to day are
Windows systems, and since Windows use HTFS it is logical that most HD
that fail would have disk or media formatted to the HTFS format.

This is the same logic that says one of the most dangerous chemical in
the US today is water, since a large number of people die every year
form overdoses.


--
2018: The year we learn to play the great game of Euchre

On Fri, 26 Apr 2019 11:35:27 -0400, Flasherly wrote:

| And last, for the grand finale -- Gold class at 2M hours MTBF.
| Definitely new territory and one I haven't looked at.
|
| WD's flagship, evidently, with these warranty specs -
|
| 2,000,000/24
| 83333.3333333/365
| 228.310502283
|
| How spectacular. You can bequeath that to five generations at an
| historical average ceded for 50 years (so used in Egyptology for
| dating its dynasties).
|
| At 2016 it would appear the largest WD as well possible at 10TB. I
| would offhand guess prices at the time of your acquisition may have
| ranged upwards of $400, if not excessively more;- with a sale
| possibly to entail at least $200US.

Nope. I only needed a 1TB working HDD (E: Editing on my main PC) and got a Gold from
Newegg for $73.

Larc

Larc wrote:
On Fri, 26 Apr 2019 11:35:27 -0400, Flasherly wrote:

| And last, for the grand finale -- Gold class at 2M hours MTBF.
| Definitely new territory and one I haven't looked at.
|
| WD's flagship, evidently, with these warranty specs -
|
| 2,000,000/24
| 83333.3333333/365
| 228.310502283
|
| How spectacular. You can bequeath that to five generations at an
| historical average ceded for 50 years (so used in Egyptology for
| dating its dynasties).
|
| At 2016 it would appear the largest WD as well possible at 10TB. I
| would offhand guess prices at the time of your acquisition may have
| ranged upwards of $400, if not excessively more;- with a sale
| possibly to entail at least $200US.

Nope. I only needed a 1TB working HDD (E: Editing on my main PC) and got a Gold from
Newegg for $73.

Larc

The small Gold (1TB/2TB/4TB) were some of the last
drives with 512N sectors. If you got a Blue today
(i.e. bottom tier), those would be mostly 512e (emulated).
The 512N is suitable for WinXP installs. The 512e would
be less suitable for WinXP (alignment) but OK for Vista+.

I have WinXP on a 512N right now, with no tricks necessary.
MSDOS partitioning, divisible by 63 numbers, makes no
difference to the performance of a 512N drive.

I try to only buy a couple disks a year, and I also try to
buy the disks from a brick and mortar store (I don't trust
how disks are bubble-wrapped in Fedex boxes). This years
purchase was 6TB Black, and the packaging seems to be the
same on those, as last years 4TB Gold. The Black is quieter
than a Black of five years ago. Doesn't buzz like one. The
Gold packaging was different enough, to suggests maybe it
was a Hitachi drive, but that probably isn't it either.
All the disks in question are air breathers, and looking
at the power numbers tends to hint at what drives are
sealed and filled with Helium.

Some WDC drives make the "dying cow" noise when the spindle
spins down. The Black I got doesn't make that noise.
Someone reported a Red Pro 10TB was making the dead cow
noise, and that was the reason that was put second on
the list when comparison shopping.

Seagate has the same tiers in their sales schemes too,
but the introduction of ****ty SMR drives in the bottom
tiers, leaves me suspicious of their other products.

How you treat your customers, actually matters... :-/

Removing critical specs from disk drive marketing
materials, won't fool everybody. It just makes the
shopping experience a miserable, drawn-out one. Lots
of "reading amateur analysis postings" to figure out
what you're actually getting for the money. The information
is out there, but it's damn hard to gather and collate.

Like, why can't all the drives have 512n, 512e, and 4Kn
notations in the datasheet ? Would that kill someone to
include the info ? I had to get a "materials for salesman"
document off a website, to have a separate list of which
ones are which. And obviously, if they indicated the
platter count on each drive, it would be "too easy" to
figure out which drives are SMR (shingled) instead
of PMR (perpendicular magnetic recording). SMR are
perpendicular, but the tracks are more closely spaced,
and have zero clearance between tracks in a 7-track cluster.
PMR recorded disks have the same spacing between all tracks.

At one time, I would have posited that the factory used
very few platter types. But the evidence now suggests
disk drives are like 9 speed bicycles, and the materials
used are different in the big drives than the small drives.
Which means the factory has more different boxes of platters
incoming, than might have been the case previously. It appears
there are still, independent suppliers of finished platters
for these drives :-) The platter companies don't seem to all
have been bought up. The motors are made by a third-party
company too. I would dearly like to know the design details
of the "dying cow" motors... :-) Just to see what they were
thinking.

Paul

On 26/04/2019 21.05, Keith Nuttle wrote:
On 4/26/2019 2:24 PM, Cows are Nice wrote:
On 04/26/2019 10:06 AM, Filip454 wrote:
W dniu 2019-04-26 o 06:31, Cows are Nice pisze:
Windows ruins hardware

Could you elaborate?

Yes. Every so often, an anti-Linux troll gets swatted for visiting
Linux groups to spread alarm that Linux will destroy their hardware.
Since we enjoy that, following up with many witty replies, I thought
it a swell idea to bring the fun to win-droids with a crosspost.

Fact - Most dead HDDs, nearly all of them, are found to be infected
with NTFS.
Linux is a minor player in the personal computer world, so statistically
that is correct.Â* The largest percentage of computers in use to day are
Windows systems, and since Windows use HTFS it is logical that most HD
that fail would have disk or media formatted to the HTFS format.

Unless you count the server farms, where Linux is prevalent ;-)

Otherwise, you are right, the stat should be percent re total disks
installed with whatever filesystem to mean something. Looking at
absolute numbers of ntfs fails means little.

But "Cows are Nice" was joking at you - he said as much :-p

--
Cheers,
Carlos E.R.

On 4/26/2019 4:45 PM, Carlos E. R. wrote:
On 26/04/2019 21.05, Keith Nuttle wrote:
On 4/26/2019 2:24 PM, Cows are Nice wrote:
On 04/26/2019 10:06 AM, Filip454 wrote:
W dniu 2019-04-26 o 06:31, Cows are Nice pisze:
Windows ruins hardware

Could you elaborate?

Yes. Every so often, an anti-Linux troll gets swatted for visiting
Linux groups to spread alarm that Linux will destroy their hardware.
Since we enjoy that, following up with many witty replies, I thought
it a swell idea to bring the fun to win-droids with a crosspost.

Fact - Most dead HDDs, nearly all of them, are found to be infected
with NTFS.
Linux is a minor player in the personal computer world, so statistically
that is correct.Â* The largest percentage of computers in use to day are
Windows systems, and since Windows use HTFS it is logical that most HD
that fail would have disk or media formatted to the HTFS format.

Unless you count the server farms, where Linux is prevalent ;-)

Otherwise, you are right, the stat should be percent re total disks
installed with whatever filesystem to mean something. Looking at
absolute numbers of ntfs fails means little.

But "Cows are Nice" was joking at you - he said as much :-p

I realized that, but that was for the linux users.

--
2018: The year we learn to play the great game of Euchre

On Fri, 26 Apr 2019 08:29:24 -0500, Charlie Hoffpauir
wrote:

Thanks for all the comments.
I was actually thinking.... I could put one or both in service as one
of my rotating backups. In that manner it would only be used once per
month, for only a few minutes. "5 days" could last months that way.
Yesterday I tried to run Seatools DOS, downloaded from the Seagate
site, but it looks like the GUI version doesn't work (on my setup at
least... it doesn't locate any drives), so I did a "long" format from
within Windows (5+ hours on the 3 TB). Results unchanged.... still
predicting 5 days. I did notice that this drive does run several
degrees hotter than any of my other drives that are in service. The
only one I've seen run hotter is a WD Black 3 TB that runs so hot I'm
afraid to use it in my system. (FYI, SSDs in my system run typically
27-28 C, HDDs about 33-35 C, and this 3TB Seagate pushes 40 C.

If you actually do experience data degradation, files that become at
worse irretrievable due to drive fault errors, you may need something
heavy duty in software to do file bit-comparisons.

Various file synchronization programs or general back-up grade
software only goes to point. Once the errors start happening and if
you really need something on a backup, say that is lost on the drive
it serves due to other than mechanical issues, then something else may
be required.*

I've retrieved, to an extent some data with this commercial program,
although if it can't do it, least to mention time factors in a
"brute" methodology, a time may have arrived for further
reconciliation, i.e. cursing the gods responsible for fating our daily
data limits to crappy high priests of engineering.

*
CDCheck is a utility for prevention, detection
and recovery of damaged files with emphasis
on error detection. It can check each your
CD/DVD (or any other media) and indicate
which files are corrupted. CDCheck reporting
features tell you exactly where the problems
are. Files on CDs, zip drives, USB keys etc.
can get damaged in a number of ways, so the
program helps you determine whether your data
is safe before it's too late. The program
also provides extremly fast binary compare
for effectively checking that file transfers
(burning, copying...) were accomplished
successfully and alerts you of differences.
Besides that CDCheck supports creation and
checking of MD5, CRC-32, SHA... hashes in
SFV, MD5 and CRC file formats. This provides
means to check backups for possible loss of
information or verify file transfers where
comparing is not possible (transfer over mail
etc.). The program can be used with all local
or removable media (CDs, DVDs, disk drives,
floppy disks, ZIP drives, USB keys. ..)
visible by the operating system (Windows
Explorer) and also with audio CDs. In
addition CDCheck also gives detailed
information (manufacturer, type, capacity...)
about inserted CD or DVD media.

Key features (details):

hash creation and checking
file/directory checking
binary file/directory compare
file/directory recovery
audio CD support
data DVD support
CD/DVD information with media ratings

On Fri, 26 Apr 2019 15:21:19 -0400, Larc
wrote:

Nope. I only needed a 1TB working HDD (E: Editing on my main PC) and got a Gold from
Newegg for $73.

Might have been after the typhoon that hit some major disc facilities
on the Pacific Rim. 1-2T drives, prior to the storm, hit their
pinnacle of value. It would be roughly five years for those same
prices to reach about what they are now: 2T for $60 (and 4T at
slightly more in post-Legacy factors).

I bought maybe four or five Samsung 1-2T drives when averaging $40-50
at those prices.

A 2T WD failure, of recent, a former Green class lasted 4 or 5 years
at relatively heavy usage. I cycled in a 2T Samsung replacement, and
paid $60, last year, for another 2T WD of similar if not "Green" kind.

Amazon, though. I like and for the most trust NewEgg, but I"m more
selective about what I order from NewEgg since they've adopted a
no-slack policy of hitting on the customer's dime in the case of an
RMA. It's not written exactly in stone, but I'm neither into whining
for sneering representative angling at odds the customer will buckle.

Anyway, it's near to a point where 2T isn't making sense at their
current pricing. I don't understand, after a couple years, what's so
hard now about the simple math of $40 for 2T and $70 for 4T. (By the
time I figure it out a 1T SSD could cost $50.)

Charlie Hoffpauir wrote:

I have a couple of Seagate Hard Drives that Hard Disk Semtinel says
are failing. One is model ST2000DM001 and the other ST3000DM008. I've
removed both from my system and recovered the data, but I'm wondering
if there's anything that can be done to "save" the drives, perhaps as
backups, or is that a useless activity and I should just toss them.
Hard Disk Sentinel says one has an expected life of 5 days, and the
other 16 days. The problems reported are bad sectors, for example, for
the 3 TB drive:
56 bad sectors
3288 bad sectors during self test
4267 errors during data transfer
power on time 587 days, 3 hours
Est remaining lifetime 5 days
total start/stop count 11,804

so, can anything be done for them, or or they scrap.

Which EDITION of HD Sentinel do you have? The payware editions include
hard disk tests. I cannot look right now because I scrapped my old
computer and am setting up a new one, so no HD Sentinel to look at yet.
From their feature comparison chart at:

https://www.hdsentinel.com/store.php

You don't get the "Complete surface analysis, refresh, reinitialise,
repair" feature set until you buy their Professional edition. That's
what I have (can see it in the Downloads folder but haven't yet got
around to installing it in the new build). I could not find in their
online manual what "refresh" and "repair" are supposed to do. My guess
is refresh will write the sector's data to elsewhere (after testing the
unused sector), exercise the sector (write different patterns), and
attempt to write the data back to the sector. If it fails, the bad
sector gets marked bad and the file table is adjusted to reflect the
data in in a different sector. Refreshing is how you eliminate "bit
rot", especially for data that is written once and thereafter rarely
accessed and usually only read; i.e., the sector never gets rewritten.
Read:

https://www.hdsentinel.com/forum/viewtopic.php?t=10769

I can't say what their repair feature does. They give a list of
features, not to which menu entry they are associated. In either case
(refresh or repair, if those are different processes), a full image
backup is recommeded before exercising either feature.

If you're not going to attempt to use software to refresh/repair the
suspect drives then store your data elsewhere, do an image backup, and
start looking to buy replacements (to which you can restore the image).

Personally I shy away from the Seagate brand and stick with WDC. Every
brand has its "flyers" (aka "outliers"): models that don't perform or
survive well. Seagate has more flyers than WDC.

On Sat, 27 Apr 2019 10:55:36 -0500, VanguardLH wrote:

Charlie Hoffpauir wrote:

I have a couple of Seagate Hard Drives that Hard Disk Semtinel says
are failing. One is model ST2000DM001 and the other ST3000DM008. I've
removed both from my system and recovered the data, but I'm wondering
if there's anything that can be done to "save" the drives, perhaps as
backups, or is that a useless activity and I should just toss them.
Hard Disk Sentinel says one has an expected life of 5 days, and the
other 16 days. The problems reported are bad sectors, for example, for
the 3 TB drive:
56 bad sectors
3288 bad sectors during self test
4267 errors during data transfer
power on time 587 days, 3 hours
Est remaining lifetime 5 days
total start/stop count 11,804

so, can anything be done for them, or or they scrap.

Which EDITION of HD Sentinel do you have? The payware editions include
hard disk tests. I cannot look right now because I scrapped my old
computer and am setting up a new one, so no HD Sentinel to look at yet.
From their feature comparison chart at:

https://www.hdsentinel.com/store.php

You don't get the "Complete surface analysis, refresh, reinitialise,
repair" feature set until you buy their Professional edition. That's
what I have (can see it in the Downloads folder but haven't yet got
around to installing it in the new build). I could not find in their
online manual what "refresh" and "repair" are supposed to do. My guess
is refresh will write the sector's data to elsewhere (after testing the
unused sector), exercise the sector (write different patterns), and
attempt to write the data back to the sector. If it fails, the bad
sector gets marked bad and the file table is adjusted to reflect the
data in in a different sector. Refreshing is how you eliminate "bit
rot", especially for data that is written once and thereafter rarely
accessed and usually only read; i.e., the sector never gets rewritten.
Read:

https://www.hdsentinel.com/forum/viewtopic.php?t=10769

I can't say what their repair feature does. They give a list of
features, not to which menu entry they are associated. In either case
(refresh or repair, if those are different processes), a full image
backup is recommeded before exercising either feature.

If you're not going to attempt to use software to refresh/repair the
suspect drives then store your data elsewhere, do an image backup, and
start looking to buy replacements (to which you can restore the image).

Personally I shy away from the Seagate brand and stick with WDC. Every
brand has its "flyers" (aka "outliers"): models that don't perform or
survive well. Seagate has more flyers than WDC.

I've had Hard disk Sentinel for longer than I can remember, but I do
remember that I upgraded it a couple of years ago. A check says it's
indeed registered to me, and the version says v4.60(7377). A quick
look at their web page says 5.40 is current. So I went through the
update process and it now shows I'm updated to 5.40.

I probably didn't make it clear enough that these two disks have been
removed from my system, I was only considering using them as backups
via my Icy Dock tray,

"Most" of my drives are Seagate, and it seems most of the problems are
with them also. But the older (and smaller) Seagates seemed to be much
better. I still have some Seagate 500 GB ST 3500418AS drives that I
used for a long time in RAID that still check out perfect. I just put
one in the Icy dock and it checks out at 834 days est, remaining life.
I use these for cloning my SSD C drive backups.