Western Digital Elements USB Disk drives

D

DJT

I have 2 WD Elements USB Disk drives. When I try to copy large files
to either of the disks the file appears to copy but the system hangs.
It appears to actually copy the file but hangs when it is renaming the
file to the correct name.

Sometimes the file is copied and an entry appears with a date which is
now not the date of the file. The System then hangs. Sometimes the
date changes to the correct one.

The only way to clear the fault is to turn off the power to the disks,
the system then continues as normal.

This appears to be on Win 7 only but not sure


Anyone have a clue how to fix this


DJT
 
P

pjp

I have 2 WD Elements USB Disk drives. When I try to copy large files
to either of the disks the file appears to copy but the system hangs.
It appears to actually copy the file but hangs when it is renaming the
file to the correct name.

Sometimes the file is copied and an entry appears with a date which is
now not the date of the file. The System then hangs. Sometimes the
date changes to the correct one.

The only way to clear the fault is to turn off the power to the disks,
the system then continues as normal.

This appears to be on Win 7 only but not sure


Anyone have a clue how to fix this


DJT
I had that problem with a WD external disk. Turned out disk went bad.
When I checked it had used up all available Relocation Sectors which
meant it had been having bad sectors for some time. It was out of
warranty and now junk (except for magnets inside it). I now check
frequently the three I have as one of them is a replacement for a
Seagate external was going bad (using up sectors but still had some
spares) that thankfully they replaced under warranty shortly before
expiry period. I suspect it would have lasted past warranty period but
not by much so ... Minute I see any of them using up Relocation Sectors
I'll check warranty well before it becomes a problem.
 
P

Paul

DJT said:
I have 2 WD Elements USB Disk drives. When I try to copy large files
to either of the disks the file appears to copy but the system hangs.
It appears to actually copy the file but hangs when it is renaming the
file to the correct name.

Sometimes the file is copied and an entry appears with a date which is
now not the date of the file. The System then hangs. Sometimes the
date changes to the correct one.

The only way to clear the fault is to turn off the power to the disks,
the system then continues as normal.

This appears to be on Win 7 only but not sure


Anyone have a clue how to fix this


DJT
The partition would have two areas in question. A large number
of free data clusters, located on the disk. And a MFT or Master
File Table. Depending on where the damaged area of the disk is,
could affect either the ability to write the data, or the
ability to update the MFT.

Even Windows could be having a problem, such as a TDSS rootkit
being present, the atapi.sys driver file hijacked and replaced
by malware. So if a system has an abnormal behavior, it might
be due to malware.

I've also had Windows NTFS become sluggish, after writing about
a terabyte of data to a disk. It got so slow, it declared a
"delayed write failure", even though there was nothing wrong
with the disk. Rebooting the system, and the resultant cleanup
of memory, fixes that.

*******

I think I'd work on a "damaged disk" theory first, as that is
more likely to have happened.

Disk diagnostics can be destructive, if for example they include
write testing. A disk diagnostic can point out bad sectors,
or the SMART statistics summary, can predict problems as well
(there are a couple statistics, if they're greater than
zero, you're in trouble).

Or, you can design your own tests (not more file copying,
which you know is not working). This utility is a port
of "disk dump", a Unix utility of long standing. The person
on this site, has made it available for Windows users.

http://www.chrysocome.net/dd

http://www.chrysocome.net/downloads/dd-0.5.zip

You use this command, to list all the disks and all the
partitions. "Partition0" means the whole disk, starting
at sector 0, will be the mode of access (i.e. wiping the
whole drive or copying the whole drive). "Partition1"
corresponds to just one partition on the disk. If it
was my C: partition, it might start at say, sector 63.
A command executed using "Partition 1" might be used
to wipe out only my C: partition (yikes!). Anyway, using
this command, shows all of the options for disk syntax,
to be used in later (potentially destructive) operations.
This command is safe, and won't hurt anything. And
can be used if you're just curious about what is on
there, sizes and so on. Sometimes I need a size value,
for some other thing I'm doing. (In Windows 7, I type
"cmd" in Start, then "Run as Administrator", to make
a command prompt with administrator privileges. Then
run the command in Command Prompt.)

dd.exe --list

That lists the syntax of the storage partitions on the disk(s).
You have to be really really careful using the following commands,
because one mistake can wipe out your system partition. (I have
actually mis-typed a command once, and once is enough.)

The size of the disk will be reported, as in

\\?\Device\Harddisk0\Partition0
Fixed hard disk media. Block size = 512
size is 500107862016

That would correspond to the first disk in Disk Management.
(The disks should be numbered in the same order as Disk Management.)

Now, I attempt to zero out the whole drive, each sector
of it. Obviously, doing this is completely destructive. Backup
the data to another disk, before you do this step. The "/dev/zero"
device, is a virtual source of sectors containing nothing but
0x00 bytes. This is a good way to erase a drive as well...

dd.exe if=/dev/zero of=\\?\Device\Harddisk0\Partition0 bs=8192

The block size specification, should be a multiple of the sector
size, which is 512 bytes in this case. The number 500,107,862,016
can be factored, and 8192 in this case, happens to divide into
that number. You want a block size larger than 512 bytes, so
the command will do I/O at a decent speed (using 512 bytes
means too many commands per second are necessary and it
runs slow).

Now, it will report "how many blocks" were written when the command
finishes. It would take me about five hours, to run that command
if my disk was in a USB2 enclosure. And at the end, it should report

61048323 blocks transferred

If I multiply 61048323 by 8192, I should get back the 500,107,862,016
original disk size (500GB). If the command "stops short", then something
is wrong.

Because of a bug in that tool, a better option is to tell it how many
blocks you expect to transfer, like this. Now the command knows
it should stop when it hits a total of 500,107,862,016 bytes of data.

dd.exe if=/dev/zero of=\\?\Device\Harddisk0\Partition0 bs=8192 count=61048323

It will report the count you requested, and how many blocks
were actually transferred. If the two values are different,
then there was a write error. Chances are, if the disk is
bad (has no reallocatable spare sectors left in the damaged area),
the command will stop short of the five hours. The block count
reported, would give you an idea of the area where the first
non-recoverable error occurred.

You can go to the Western Digital site, and download their diagnostic
for your disk. But what it will display, is an "error code". And
the value of the error code, is what you use to convince them
they need to provide warranty service on the drive (if there
is any warranty time left). So to them, the main value of
the diagnostic, is to provide justification for usage of the
warranty. And not, to provide plain English explanations of
what is wrong with the disk.

There are other tools, such as Spinrite, a commercial tool that
claims to be able to repair disks by getting them to reallocate
sectors or whatever. If you believe in such things, you can
try that on a flaky disk. As far as I know, on an IDE drive,
there is no easy way to reset the reallocated sectors, and
have the drive work out the failed sectors all over again.
Reallocation on IDE is "automatic", and if a sector was
reallocated and a spare sector used "by mistake", you can't
easily tell the disk to "do over". When I used to own and work
on SCSI disks, they have an option to start over again. Which
I would use occasionally, just for "fun", to see what would
happen. The disk usually locates the same bad sectors again,
and spares them out. And doing so, just slows the disk down.

If your own, home grown test, detects a problem, or you
examine the SMART stats and see trouble, then it's time
to retire the drive, and start using the backup you just
made. Assuming that is, that the backup operation actually
succeeded error free. In any case, any time you're having
disk trouble, the very first thing to do, is acquire a
spare disk (or two) for your recovery, repair, and test
operations.

Hard drives in uncooled USB enclosures, aren't such a good deal.
My USB enclosure has a 40mm fan, which is close to useless. I
have some enclosures here, where they came with a 40mm fan,
but no cooling intake vents! There are some stupid designers
in this world. (I had to drill holes in the bottom of
that enclosure, near the front.) The best drives for cooling,
have a slow speed 80mm fan on a top or bottom surface, but I
don't own one of those. In order to control exhaust heat, the
firmware in the hard drive, will support an aggressive
"spin down" behavior, such that the drive will spin down
during inactive moments. This can result in the heads being
parked 400 to 800 times per day. While the very latest disks
are rated for 300,000 head parks, this is a silly way to burn
them up. By parking the heads, the drive electrical power
requirement drops significantly, and then the drive will
cool off. It's a piss poor way to control heat. And can't
be used, when you copy data for two hours to the disk, and
the heat just builds up. (The heads can't park if you're
doing continuous writes.)

If you need a temperature readout, you can leave HDTune
running while you're doing your normal file copy. Then
see what kind of temperature results. Above 50C, I'd be
just a bit concerned. This version has a bit of trouble
with graphics rendering, so I don't know how well it'll
work in Windows 7. This copy is free. If you get an
error, the command may need to be "elevated" (Run as
Administrator), in order to do disk I/O.

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

Running CHKDSK on the disk, is a way to test for
structural errors as well. But could also take a while.

Paul
 
D

DJT

I had that problem with a WD external disk. Turned out disk went bad.
When I checked it had used up all available Relocation Sectors which
meant it had been having bad sectors for some time. It was out of
warranty and now junk (except for magnets inside it). I now check
frequently the three I have as one of them is a replacement for a
Seagate external was going bad (using up sectors but still had some
spares) that thankfully they replaced under warranty shortly before
expiry period. I suspect it would have lasted past warranty period but
not by much so ... Minute I see any of them using up Relocation Sectors
I'll check warranty well before it becomes a problem.
I had previously run diagnostics on these disks and they passed.

I ran them again and they both passed again

The problem only appears to occur when updating the directories after
copying the file.


DJT
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top