Delayed Write Failed


T

Tom

Hello.

I own a P4C800-E Deluxe motherboard, with Windows XP SP2 installed. I
realize it's an old motherboard, but I can't afford to purchase a new
computer. I'm hoping to get a few more years out of this one. :)

The issue has to do with a little over one year old Rosewill 'R2-JBOD'
external enclosure (self powered). The Rosewill is configured as JBOD,
with two Western Digital WD1002FBYS-02A6B0 1TB drives installed. The
Rosewill is connected directly to one of the motherboard USB ports.
Windows has the drives configured for quick removal and I always use the
'Safely Remove Hardware' to dismount the device. The Rosewill has worked
flawlessly for about a year in it's present configuration.

About one month ago, I went to defrag the Rosewill drives and several
minutes into the defrag I receive a pop-up that said there was a
'Delayed Write Failed', that data was "lost". I tried various things,
but in the end what I thought solved the issue was a thorough cleaning
of the inside of the computer case and the Rosewill enclosure. The issue
immediately stopped after the cleaning. At the time I used the drives
for approximately a couple more days without issue and the Rosewill was
then powered down.

One week ago I needed the Rosewill drives, and used them continuously
for four (4) days solid, until I needed to reboot the computer. After
rebooting, I decided to defrag the drives and I once again received a
'Delayed Write Failed'. I have been defragging the system and USB drives
with PerfectDisk for years without any issues.

Typically when there is a 'Delayed Write Failed' it says which folder is
"lost" and if I check that folder, all the files are gone. However, if I
reboot the computer, all the files are back and they are completely
intact. Which folder and which files are missing, change as I reboot the
computer. It's as if whatever is causing the issue, at that specific
point in time, is what's "lost". Also, most of the time I can hear the
Windows drive un-mount chime and immediately the mount chine -- like the
drive is momentarily dropping and coming back.

I've used CHKDSK when the 'Delayed Write Failed' has occurred and
especially when the files are supposedly missing and it reports that
there is nothing wrong with the drive. It seems whenever I use CHKDSK
nothing is ever found. Maybe I need better hard drive checking software,
what would you recommend?

The only other odd thing to happen recently was during one of the
reboots I received a BSOD. The error was 'UNMOUNTABLE_BOOT_VOLUME'
(Stop: 0x000000ED). According to Microsoft they say that this error
message occurs "When Volume on IDE Drive with Caching [is] Enabled".
They also say that this issue was fixed with Windows XP SP1, I have SP2.
I checked the boot drive with CHKDSK and CHKDSK again reports no errors.
And, it could be a coincidence, since in all these years it has only
happened twice, with a fair amount of time passing by and with several
boots in-between.

Some other things I've tried include:

1) Moving the USB cable for the Rosewill to one of the other seven
motherboard USB ports.

2) I've tried a different USB cable that's currently used for a
different brand of USB enclosure, which works fine with that drive.

3) I've removed the outer shell of the Rosewill to allow better air flow
and the metal hard drive mounting frame is only warm to the touch.

4) Rosewill fan is enabled.

5) I've cycled the USB connector into the Rosewill and motherboard ports
hoping to clear any possible oxidation. In addition, made sure that
the BIG and JBOD mode switch inside the Rosewill was firmly set to
JBOD by cycling the switch back and forth.

Other than these two issues, the computer is working fine. I don't know
if it's a motherboard (USB) issue or a enclosure (hard drive) issue.
Hence, posting to two Newsgroups, with followup set. Sorry for the
length of this post, I was trying to be as thorough as I could. If I
missed any important details, please let me know.

Thank you.
 
Ad

Advertisements

A

Arno

[sorry for top-posting]

This might be some diconnect issue with the USB bus. The
umount/remount sounds like what you see when you pull
the dvice and immediately re-plug. Setting to fast
removal only does so much and on longer writes (which
you get during defragging), you can get disconnects in
the middle of writes. The good news is that, if the defragger
is programmed well, it will copy data first and then
adjust the metadata (directory, etc.), which gives you
the impression that nothing was lost. In fact, the new copy
of the data written was lost, but the old one is still there
and the metadata still points to it. This is the likely
outcome, but real data-loss is also possble.

I suspect a faulty connection somewhere in the cables.
Cleaning the inside of the enclosure can, like shaking
or (no joke) screaming at the thing, temporarily fix
the faulty connector, as they are very sensitive to
any kind of vibration or movement. It is not permanent
though. I strongly advise you to not trust this enclosure
until you find and fix the fault or replace the enclosure.

Arno



In comp.sys.ibm.pc.hardware.storage Tom said:
I own a P4C800-E Deluxe motherboard, with Windows XP SP2 installed. I
realize it's an old motherboard, but I can't afford to purchase a new
computer. I'm hoping to get a few more years out of this one. :)
The issue has to do with a little over one year old Rosewill 'R2-JBOD'
external enclosure (self powered). The Rosewill is configured as JBOD,
with two Western Digital WD1002FBYS-02A6B0 1TB drives installed. The
Rosewill is connected directly to one of the motherboard USB ports.
Windows has the drives configured for quick removal and I always use the
'Safely Remove Hardware' to dismount the device. The Rosewill has worked
flawlessly for about a year in it's present configuration.
About one month ago, I went to defrag the Rosewill drives and several
minutes into the defrag I receive a pop-up that said there was a
'Delayed Write Failed', that data was "lost". I tried various things,
but in the end what I thought solved the issue was a thorough cleaning
of the inside of the computer case and the Rosewill enclosure. The issue
immediately stopped after the cleaning. At the time I used the drives
for approximately a couple more days without issue and the Rosewill was
then powered down.
One week ago I needed the Rosewill drives, and used them continuously
for four (4) days solid, until I needed to reboot the computer. After
rebooting, I decided to defrag the drives and I once again received a
'Delayed Write Failed'. I have been defragging the system and USB drives
with PerfectDisk for years without any issues.
Typically when there is a 'Delayed Write Failed' it says which folder is
"lost" and if I check that folder, all the files are gone. However, if I
reboot the computer, all the files are back and they are completely
intact. Which folder and which files are missing, change as I reboot the
computer. It's as if whatever is causing the issue, at that specific
point in time, is what's "lost". Also, most of the time I can hear the
Windows drive un-mount chime and immediately the mount chine -- like the
drive is momentarily dropping and coming back.
I've used CHKDSK when the 'Delayed Write Failed' has occurred and
especially when the files are supposedly missing and it reports that
there is nothing wrong with the drive. It seems whenever I use CHKDSK
nothing is ever found. Maybe I need better hard drive checking software,
what would you recommend?
The only other odd thing to happen recently was during one of the
reboots I received a BSOD. The error was 'UNMOUNTABLE_BOOT_VOLUME'
(Stop: 0x000000ED). According to Microsoft they say that this error
message occurs "When Volume on IDE Drive with Caching [is] Enabled".
They also say that this issue was fixed with Windows XP SP1, I have SP2.
I checked the boot drive with CHKDSK and CHKDSK again reports no errors.
And, it could be a coincidence, since in all these years it has only
happened twice, with a fair amount of time passing by and with several
boots in-between.
Some other things I've tried include:
1) Moving the USB cable for the Rosewill to one of the other seven
motherboard USB ports.
2) I've tried a different USB cable that's currently used for a
different brand of USB enclosure, which works fine with that drive.
3) I've removed the outer shell of the Rosewill to allow better air flow
and the metal hard drive mounting frame is only warm to the touch.
4) Rosewill fan is enabled.
5) I've cycled the USB connector into the Rosewill and motherboard ports
hoping to clear any possible oxidation. In addition, made sure that
the BIG and JBOD mode switch inside the Rosewill was firmly set to
JBOD by cycling the switch back and forth.
 
R

Rod Speed

Tom said:
I own a P4C800-E Deluxe motherboard, with Windows XP SP2 installed.
I realize it's an old motherboard, but I can't afford to purchase a new
computer. I'm hoping to get a few more years out of this one. :)

It isnt that old and should continue to work as well as it ever did.
The issue has to do with a little over one year old Rosewill 'R2-JBOD'
external enclosure (self powered). The Rosewill is configured as JBOD,
with two Western Digital WD1002FBYS-02A6B0 1TB drives installed.
The Rosewill is connected directly to one of the motherboard USB ports.
Windows has the drives configured for quick removal and I always use
the 'Safely Remove Hardware' to dismount the device. The Rosewill
has worked flawlessly for about a year in it's present configuration.
About one month ago, I went to defrag the Rosewill drives and
several minutes into the defrag I receive a pop-up that said there
was a 'Delayed Write Failed', that data was "lost".

That last isnt necessarily correct with a defrag, a well written one
should only adjust the directory entrys after a write has succeeded.
You dont lose anything when the write fails when the directory
entrys still list the old location for the data that was being moved.
I tried various things, but in the end what I thought solved the issue
was a thorough cleaning of the inside of the computer case and the
Rosewill enclosure. The issue immediately stopped after the cleaning.

That is absolutely guaranteed to be a coincidence.
Dirty cases cant cause delayed write failures.
At the time I used the drives for approximately a couple more days without issue

Because its an intermittent fault.
and the Rosewill was then powered down.
One week ago I needed the Rosewill drives, and used them
continuously for four (4) days solid, until I needed to reboot
the computer. After rebooting, I decided to defrag the drives

There is no point in furiously defragging drives at anything like that rate.
and I once again received a 'Delayed Write Failed'. I have been defragging
the system and USB drives with PerfectDisk for years without any issues.

Because there was no fault at that time.

Defragging works the drives a lot harder than anything else does.
Typically when there is a 'Delayed Write Failed' it says which
folder is "lost" and if I check that folder, all the files are gone.
However, if I reboot the computer, all the files are back and they
are completely intact.

Because the defragger is written properly.
Which folder and which files are missing, change as I reboot
the computer. It's as if whatever is causing the issue, at that
specific point in time, is what's "lost".

Nope, thats an illusion.
Also, most of the time I can hear the Windows drive
un-mount chime and immediately the mount chine --
like the drive is momentarily dropping and coming back.

Thats the fault, the drive is going offline for some reason
and thats whats producing the delayed write failure.
I've used CHKDSK when the 'Delayed Write Failed' has occurred
and especially when the files are supposedly missing and it reports
that there is nothing wrong with the drive.

For the reason I listed at the top.
It seems whenever I use CHKDSK nothing is ever found.
Maybe I need better hard drive checking software,

Nope, the drive contents are fine.
what would you recommend?

CHKDSK is fine as long as it doesnt hang. It can do
that very occassionally with some drive corruption.
The only other odd thing to happen recently was during one of the reboots
I received a BSOD. The error was 'UNMOUNTABLE_BOOT_VOLUME'
(Stop: 0x000000ED). According to Microsoft they say that this error
message occurs "When Volume on IDE Drive with Caching [is] Enabled".
They also say that this issue was fixed with Windows XP SP1, I have SP2.

So the problem they are talking about isnt your problem.

Your problem is that the drive sometimes goes offline and you can hear that happening.
I checked the boot drive with CHKDSK and CHKDSK again
reports no errors. And, it could be a coincidence, since in
all these years it has only happened twice, with a fair amount
of time passing by and with several boots in-between.

Likely the fault has been around for some time and has lately got much worse.
Some other things I've tried include:
1) Moving the USB cable for the Rosewill to one of the other seven
motherboard USB ports.

You need to do more than just do that once when
its an intermittent fault that doesnt show up very
often to prove anything about the port being used.

You need to move the drive to another port and leave it there
and see if the delayed write failure is ever seen again.
2) I've tried a different USB cable that's currently used for a
different brand of USB enclosure, which works fine with that drive.

Ditto with the above.
3) I've removed the outer shell of the Rosewill to allow better air flow
and the metal hard drive mounting frame is only warm to the touch.

It wont be that given it worked fine for years.
4) Rosewill fan is enabled.
5) I've cycled the USB connector into the Rosewill and
motherboard ports hoping to clear any possible oxidation.

Its better to just put it on a different port with a different cable and
use it like that and see if the delayed write failure shows up again.
In addition, made sure that the BIG and JBOD mode switch inside the
Rosewill was firmly set to JBOD by cycling the switch back and forth.
Other than these two issues, the computer is working fine. I don't know
if it's a motherboard (USB) issue or a enclosure (hard drive) issue.

The way to distinguish between those two possibilitys is to move
the drive to another USB port with a different cable and leave it
there and see if the delayed write fault ever shows up again or
not. If it does, its likely the Rosewell enclosure going bad.
Hence, posting to two Newsgroups,

Thats fine.
with followup set.

Thats not, I have put them back to the original again so others
who are interested can see the reply.
Sorry for the length of this post, I was trying to be as thorough as I could.

Thats much better than not posting all the detail of what you have tried etc.
If I missed any important details, please let me know.

Its pretty good, although we obviously cant know what vital
info on what you might have tried you havent mentioned.
 
T

Tom

In said:
This might be some disconnect issue with the USB bus. The
unmount/remount sounds like what you see when you pull the device and
immediately re-plug.

Yep, two chimes in quick succession.
Setting to fast removal only does so much and on longer writes (which
you get during defragging), you can get disconnects in the middle of
writes.

Which use to work fine until recently. Like I mentioned, the Rosewill
worked flawlessly for about a year in it's present configuration. Then
one day out of the blue I started to get Delayed Write Failed errors
during a routine defrag.
The good news is that, if the defragger is programmed well, it will
copy data first and then adjust the metadata (directory, etc.), which
gives you the impression that nothing was lost. In fact, the new copy
of the data written was lost, but the old one is still there and the
metadata still points to it. This is the likely outcome, but real
data-loss is also possible.

I hope PerfectDisk is one such program like you described, it was highly
recommended. I have not noticed actual data loss, yet. The data that the
Delayed Write Failed reported as "lost", so far has returned after a
reboot.
I suspect a faulty connection somewhere in the cables. Cleaning the
inside of the enclosure can, like shaking or (no joke) screaming at
the thing, temporarily fix the faulty connector, as they are very
sensitive to any kind of vibration or movement. It is not permanent
though. I strongly advise you to not trust this enclosure until you
find and fix the fault or replace the enclosure.

Speaking of the Rosewill enclosure. I was reading some older posts and I
think it might be important to explain what I mean by JBOD. I'm using
the definition that Rosewill uses for the R2-JBOD enclosure. On the
Rosewill website they say:

JBOD: It enables each hard drive to be seen separately as single
drives, showing 2 HDD capacities.

http://www.rosewill.com/products/1095/productDetail.htm

In some of the earlier posts I read that JBOD was described as placing
one drives available free space after another. Just to be clear, the two
1TB drives that are installed in the Rosewill show up in Windows as two
separate drives (M: and N:). I went with JBOD mode over BIG mode to
allow for individual failure of the hard drives. Two hard drives in a
single enclosure for convenience. Just wanted to clear that up.

There are a couple of things I thought of trying after posting, that
I'll do tomorrow. One would be to remove one of the hard drives from the
Rosewill and see if the issue remains. By removing one of the hard
drives it would reduce the load on the power supply. Perhaps
PerfectDisk's defrag is showing a weakness in the power brick.

The other thing I'll try is to take out the offending hard drive and
connect it via USB with a Vantec hard drive adapter. This is a little
device that has a USB plug at one end of the cable and a 2.5/3.5
IDE/SATA interface at the other end. It has it's own power supply and
it's main purpose is to connect spare hard drives via USB without the
need for an enclosure. I'd like to see how the hard drive behaves
outside of the Rosewill.

Thank you for the reply.
 
M

Mike Tomlinson

Tom said:
Which use to work fine until recently. Like I mentioned, the Rosewill
worked flawlessly for about a year in it's present configuration. Then
one day out of the blue I started to get Delayed Write Failed errors
during a routine defrag.

Defragging is quite hard on the drive and it will be drawing peak
current. I wonder if the power supply to the Rosewill is starting to
fail (bad caps?) Might be worth checking the voltage at the drive power
connector while it is operating.

I don't think you said whether you get the delayed write errors on both
drives, but if you do, there has to be a common factor. You've tried a
different motherboard USB port and a known good cable.
There are a couple of things I thought of trying after posting, that
I'll do tomorrow. One would be to remove one of the hard drives from the
Rosewill and see if the issue remains. By removing one of the hard
drives it would reduce the load on the power supply. Perhaps
PerfectDisk's defrag is showing a weakness in the power brick.

That'd be a good thing to try.
 
G

GMAN

Hello.

I own a P4C800-E Deluxe motherboard, with Windows XP SP2 installed. I
realize it's an old motherboard, but I can't afford to purchase a new
computer. I'm hoping to get a few more years out of this one. :)

The issue has to do with a little over one year old Rosewill 'R2-JBOD'
external enclosure (self powered). The Rosewill is configured as JBOD,
with two Western Digital WD1002FBYS-02A6B0 1TB drives installed. The
Rosewill is connected directly to one of the motherboard USB ports.
Windows has the drives configured for quick removal and I always use the
'Safely Remove Hardware' to dismount the device. The Rosewill has worked
flawlessly for about a year in it's present configuration.

About one month ago, I went to defrag the Rosewill drives and several
minutes into the defrag I receive a pop-up that said there was a
'Delayed Write Failed', that data was "lost". I tried various things,
but in the end what I thought solved the issue was a thorough cleaning
of the inside of the computer case and the Rosewill enclosure. The issue
immediately stopped after the cleaning. At the time I used the drives
for approximately a couple more days without issue and the Rosewill was
then powered down.

One week ago I needed the Rosewill drives, and used them continuously
for four (4) days solid, until I needed to reboot the computer. After
rebooting, I decided to defrag the drives and I once again received a
'Delayed Write Failed'. I have been defragging the system and USB drives
with PerfectDisk for years without any issues.

Typically when there is a 'Delayed Write Failed' it says which folder is
"lost" and if I check that folder, all the files are gone. However, if I
reboot the computer, all the files are back and they are completely
intact. Which folder and which files are missing, change as I reboot the
computer. It's as if whatever is causing the issue, at that specific
point in time, is what's "lost". Also, most of the time I can hear the
Windows drive un-mount chime and immediately the mount chine -- like the
drive is momentarily dropping and coming back.

I've used CHKDSK when the 'Delayed Write Failed' has occurred and
especially when the files are supposedly missing and it reports that
there is nothing wrong with the drive. It seems whenever I use CHKDSK
nothing is ever found. Maybe I need better hard drive checking software,
what would you recommend?

The only other odd thing to happen recently was during one of the
reboots I received a BSOD. The error was 'UNMOUNTABLE_BOOT_VOLUME'
(Stop: 0x000000ED). According to Microsoft they say that this error
message occurs "When Volume on IDE Drive with Caching [is] Enabled".
They also say that this issue was fixed with Windows XP SP1, I have SP2.
I checked the boot drive with CHKDSK and CHKDSK again reports no errors.
And, it could be a coincidence, since in all these years it has only
happened twice, with a fair amount of time passing by and with several
boots in-between.

Some other things I've tried include:

1) Moving the USB cable for the Rosewill to one of the other seven
motherboard USB ports.

2) I've tried a different USB cable that's currently used for a
different brand of USB enclosure, which works fine with that drive.

3) I've removed the outer shell of the Rosewill to allow better air flow
and the metal hard drive mounting frame is only warm to the touch.

4) Rosewill fan is enabled.

5) I've cycled the USB connector into the Rosewill and motherboard ports
hoping to clear any possible oxidation. In addition, made sure that
the BIG and JBOD mode switch inside the Rosewill was firmly set to
JBOD by cycling the switch back and forth.

Other than these two issues, the computer is working fine. I don't know
if it's a motherboard (USB) issue or a enclosure (hard drive) issue.
Hence, posting to two Newsgroups, with followup set. Sorry for the
length of this post, I was trying to be as thorough as I could. If I
missed any important details, please let me know.

Thank you.

Disable write caching on both drives.
 
Ad

Advertisements

R

Rod Speed

GMAN said:
Tom said:
Hello.

I own a P4C800-E Deluxe motherboard, with Windows XP SP2 installed. I
realize it's an old motherboard, but I can't afford to purchase a new
computer. I'm hoping to get a few more years out of this one. :)

The issue has to do with a little over one year old Rosewill
'R2-JBOD' external enclosure (self powered). The Rosewill is
configured as JBOD, with two Western Digital WD1002FBYS-02A6B0 1TB
drives installed. The Rosewill is connected directly to one of the
motherboard USB ports. Windows has the drives configured for quick
removal and I always use the 'Safely Remove Hardware' to dismount
the device. The Rosewill has worked flawlessly for about a year in
it's present configuration.

About one month ago, I went to defrag the Rosewill drives and several
minutes into the defrag I receive a pop-up that said there was a
'Delayed Write Failed', that data was "lost". I tried various things,
but in the end what I thought solved the issue was a thorough
cleaning of the inside of the computer case and the Rosewill
enclosure. The issue immediately stopped after the cleaning. At the
time I used the drives for approximately a couple more days without
issue and the Rosewill was then powered down.

One week ago I needed the Rosewill drives, and used them continuously
for four (4) days solid, until I needed to reboot the computer. After
rebooting, I decided to defrag the drives and I once again received a
'Delayed Write Failed'. I have been defragging the system and USB
drives with PerfectDisk for years without any issues.

Typically when there is a 'Delayed Write Failed' it says which
folder is "lost" and if I check that folder, all the files are gone.
However, if I reboot the computer, all the files are back and they
are completely intact. Which folder and which files are missing,
change as I reboot the computer. It's as if whatever is causing the
issue, at that specific point in time, is what's "lost". Also, most
of the time I can hear the Windows drive un-mount chime and
immediately the mount chine -- like the drive is momentarily
dropping and coming back.

I've used CHKDSK when the 'Delayed Write Failed' has occurred and
especially when the files are supposedly missing and it reports that
there is nothing wrong with the drive. It seems whenever I use CHKDSK
nothing is ever found. Maybe I need better hard drive checking
software, what would you recommend?

The only other odd thing to happen recently was during one of the
reboots I received a BSOD. The error was 'UNMOUNTABLE_BOOT_VOLUME'
(Stop: 0x000000ED). According to Microsoft they say that this error
message occurs "When Volume on IDE Drive with Caching [is] Enabled".
They also say that this issue was fixed with Windows XP SP1, I have
SP2. I checked the boot drive with CHKDSK and CHKDSK again reports
no errors. And, it could be a coincidence, since in all these years
it has only happened twice, with a fair amount of time passing by
and with several boots in-between.

Some other things I've tried include:

1) Moving the USB cable for the Rosewill to one of the other seven
motherboard USB ports.

2) I've tried a different USB cable that's currently used for a
different brand of USB enclosure, which works fine with that drive.

3) I've removed the outer shell of the Rosewill to allow better air
flow and the metal hard drive mounting frame is only warm to the
touch.

4) Rosewill fan is enabled.

5) I've cycled the USB connector into the Rosewill and motherboard
ports hoping to clear any possible oxidation. In addition, made
sure that the BIG and JBOD mode switch inside the Rosewill was
firmly set to JBOD by cycling the switch back and forth.

Other than these two issues, the computer is working fine. I don't
know if it's a motherboard (USB) issue or a enclosure (hard drive)
issue. Hence, posting to two Newsgroups, with followup set. Sorry
for the length of this post, I was trying to be as thorough as I
could. If I missed any important details, please let me know.

Thank you.

Disable write caching on both drives.

Cant be that, it worked fine for years.
 
T

Tom

Hello.

Sorry for the delay. Life's unexpected tasks hasn't allowed me to run
further tests and report the findings until now. Thank you goes out to
all who have provided responses. I appreciate your interest in this
issue. I've not set a followup per Rod's suggestion. All tests are
performed after a reboot.

First, for those that do not know, the 8 USB ports on the P4C800-E
Deluxe motherboard fall into two separate clusters of ports. There are
the four USB ports that are directly attached to the motherboard's rear
panel, which I will call for the purpose of this post the primary USB
ports. Then there are an additional four USB ports on a ASUS supplied
plate that gets attached to one of the expansion slots on the back of
the case and with two approximately 8 inch cables gets attached to two
headers on the motherboard, which I will call the secondary USB ports.

Also worth noting at this point the Rosewill has both M: and N: hard
drives installed and is attached to one of the primary USB ports. The
other three primary USB ports are vacant. Attached to the secondary USB
ports with 3 meter cables are the communication between the computer and
a UPS, printer, Mouse and the monitor. A quick note about the monitor.
Built into the stand of the monitor is a two USB port hub. I often use
the monitor's USB ports as a convenient way to attach Flash drives.

The first test I ran with the above USB configuration was Western
Digital's Data Lifeguard Diagnostics software. According to WinDLG, as
expected both hard drives passed.

Next, from Device Manager I un-installed the four USB Universal Host
Controllers and the one USB2 Enhanced Host Controller. Essentially this
removes the entire USB branch from Device Manager. After a reboot the
hardware was immediately discovered by Windows without issue. Ran
PerfectDisk and 6 minutes later received a Delayed Write Failed.

Next, I removed the N: hard drive from the Rosewill enclosure and then
attached the hard drive to the Vantec adapter (this device was described
in my previous reply). Plugged the Vantec's USB cable into the side of
the monitor. Re-ran PerfectDisk and 45 minutes later without finishing
the defrag, there was no Delayed Write Failed.

Next, to see if the previous test was a fluke, I put the N: hard drive
back into the Rosewill enclosure. Re-ran PerfectDisk and several minutes
later received a Delayed Write Failed.

Next, I swapped the location of the M: and N: hard drive inside the
Rosewill enclosure to see if there was a SATA/Power connector issue.
Re-ran PerfectDisk and several minutes later received a Delayed Write
Failed.

Next, to test the power supply of the Rosewill enclosure, I removed the
M: hard drive, leaving the N: hard drive. Re-ran PerfectDisk and several
minutes later received a Delayed Write Failed.

Next, with only the N: hard drive still in the Rosewill enclosure, I
moved the USB cable back to the monitor USB port. Re-ran PerfectDisk and
9 hours later PerfectDisk completed without issue.

Next, I decided to move the four USB devices from the secondary USB
ports and connect them to the primary USB ports. Connected the Rosewill
with the single N: hard drive to a secondary USB port. Re-ran
PerfectDisk and 5 hours later PerfectDisk completed without issue.

Next, I put back into the Rosewill enclosure the M: hard drive (now both
hard drives are installed) and the Rosewill enclosure is still connected
to a secondary USB port. Re-ran PerfectDisk and multiple hours later
PerfectDisk completed without issue.

Although I could be wrong, it seems to me that the issue with Delayed
Write Failed isn't with the hard drives, the Rosewill's 3 meter USB
cable (6 meters if you count the monitor's cable), the Rosewill or the
Rosewill's power brick. That the issue is more likely with the primary
USB ports. I can't explain why, when I earlier tested the secondary USB
ports, that there was a Delayed Write Failed, other than to say I pulled
one cable out at a time. Perhaps a IRQ conflict? I was extremely tired,
so maybe I dreamed the entire sequence. :)

Anyway, at this point I can't get a Delayed Write Failed to happen as
long as the Rosewill sits on a secondary USB port. I've run a couple
other tests, such as running a MD5 check, which took hours to complete,
and there was no Delayed Write Failed. It's obvious that something has
changed with the primary USB ports and not for the better. What do you
think?

Thank you.
 
T

Tom

In said:
Next, to see if the previous test was a fluke, I put the N: hard drive
back into the Rosewill enclosure. Re-ran PerfectDisk and several
minutes later received a Delayed Write Failed.

This should read:

Next, to see if the previous test was a fluke, I put the N: hard
drive back into the Rosewill enclosure and back on the primary
USB port. Re-ran PerfectDisk and several minutes later received a
Delayed Write Failed.

Also, the Rosewill USB cable remained on the primary USB port through
the next couple of tests, until as noted, was moved back to the
monitor's USB port.

Sorry about the confusion.
 
R

Rod Speed

Tom wrote
Sorry for the delay. Life's unexpected tasks hasn't allowed me to run
further tests and report the findings until now. Thank you goes out to
all who have provided responses. I appreciate your interest in this
issue. I've not set a followup per Rod's suggestion. All tests are
performed after a reboot.
First, for those that do not know, the 8 USB ports on the P4C800-E
Deluxe motherboard fall into two separate clusters of ports. There are
the four USB ports that are directly attached to the motherboard's
rear panel, which I will call for the purpose of this post the
primary USB ports. Then there are an additional four USB ports on a
ASUS supplied plate that gets attached to one of the expansion slots
on the back of the case and with two approximately 8 inch cables gets
attached to two headers on the motherboard, which I will call the
secondary USB ports.
Also worth noting at this point the Rosewill has both M: and N: hard
drives installed and is attached to one of the primary USB ports. The
other three primary USB ports are vacant. Attached to the secondary
USB ports with 3 meter cables are the communication between the
computer and a UPS, printer, Mouse and the monitor. A quick note
about the monitor. Built into the stand of the monitor is a two USB
port hub. I often use the monitor's USB ports as a convenient way to
attach Flash drives.
The first test I ran with the above USB configuration was Western
Digital's Data Lifeguard Diagnostics software. According to WinDLG,
as expected both hard drives passed.
Next, from Device Manager I un-installed the four USB Universal Host
Controllers and the one USB2 Enhanced Host Controller. Essentially
this removes the entire USB branch from Device Manager. After a
reboot the hardware was immediately discovered by Windows without
issue. Ran PerfectDisk and 6 minutes later received a Delayed Write Failed.
Next, I removed the N: hard drive from the Rosewill enclosure and then
attached the hard drive to the Vantec adapter (this device was
described in my previous reply). Plugged the Vantec's USB cable into
the side of the monitor. Re-ran PerfectDisk and 45 minutes later
without finishing the defrag, there was no Delayed Write Failed.
Next, to see if the previous test was a fluke, I put the N: hard drive
back into the Rosewill enclosure. Re-ran PerfectDisk and several
minutes later received a Delayed Write Failed.
Next, I swapped the location of the M: and N: hard drive inside the
Rosewill enclosure to see if there was a SATA/Power connector issue.
Re-ran PerfectDisk and several minutes later received a Delayed Write
Failed.
Next, to test the power supply of the Rosewill enclosure, I removed
the M: hard drive, leaving the N: hard drive. Re-ran PerfectDisk and
several minutes later received a Delayed Write Failed.
Next, with only the N: hard drive still in the Rosewill enclosure,
I moved the USB cable back to the monitor USB port. Re-ran
PerfectDisk and 9 hours later PerfectDisk completed without issue.
Next, I decided to move the four USB devices from the secondary USB
ports and connect them to the primary USB ports. Connected the
Rosewill with the single N: hard drive to a secondary USB port. Re-ran
PerfectDisk and 5 hours later PerfectDisk completed without issue.
Next, I put back into the Rosewill enclosure the M: hard drive (now
both hard drives are installed) and the Rosewill enclosure is still
connected to a secondary USB port. Re-ran PerfectDisk and multiple
hours later PerfectDisk completed without issue.
Although I could be wrong, it seems to me that the issue with Delayed
Write Failed isn't with the hard drives, the Rosewill's 3 meter USB
cable (6 meters if you count the monitor's cable), the Rosewill or the
Rosewill's power brick.

You cant be sure of that, because the fault is clearly intermittent and went
away for quite a while just by cleaning the Rosewill and system case.
That the issue is more likely with the primary USB ports.

You cant be sure of that either given that its clearly an intermittent fault
that can go away for days.
I can't explain why, when I earlier tested the secondary USB
ports, that there was a Delayed Write Failed, other than to say
I pulled one cable out at a time. Perhaps a IRQ conflict?

Much more likely to just be an intermittent fault fooling you.
I was extremely tired, so maybe I dreamed the entire sequence. :)
Anyway, at this point I can't get a Delayed Write Failed to
happen as long as the Rosewill sits on a secondary USB port.

You have seen a problem there in the past tho.
I've run a couple other tests, such as running a MD5 check,
which took hours to complete, and there was no Delayed
Write Failed. It's obvious that something has changed with
the primary USB ports and not for the better.

You cant say that given that you have seen a delayed write failure
on a secondary port and that you clearly have an intermittent fault.
What do you think?

That you really need to try something very basic like changing the USB cable
and see if you ever get any delayed write failures with a different one. And it
need to work fine for a week or more before you can conclude thats where
the problem lay. If you get a delayed write failure with a different USB cable
you then need to use the Vantec adapter for weeks or until a delayed write
failure shows up with that, to eliminate the Rosewell enclosure being the problem.

You have already shown that you do get delayed write failures
on both what you call the primary and secondary USB ports.
 
M

Mike Tomlinson

Tom said:
Although I could be wrong, it seems to me that the issue with Delayed
Write Failed isn't with the hard drives, the Rosewill's 3 meter USB >
cable (6 meters if you count the monitor's cable),

I'd be very reluctant to use even a 3m cable with a storage device, and
6m, even with the hub in the monitor, is too long.

Try a *good quality* shorter cable - 1m - on the primary USB port and
see how that goes.
 
Ad

Advertisements

T

Tom

In said:
I'd be very reluctant to use even a 3m cable with a storage device,
and 6m, even with the hub in the monitor, is too long.

Try a *good quality* shorter cable - 1m - on the primary USB port and
see how that goes.

Ack, that's a typo. I was attempting to convert 3 and 6 feet length
cables to meters in my head and inadvertently combined both without the
calculation.

That should read:

Although I could be wrong, it seems to me that the issue with
Delayed Write Failed isn't with the hard drives, the Rosewill's 1
meter USB cable (2 meters if you count the monitor's cable),

Sorry about the confusion.
 
T

Tom

In you wrote:

[My Typo Corrected]
You cant be sure of that, because the fault is clearly intermittent
and went away for quite a while just by cleaning the Rosewill and
system case.

True, it did go away for awhile after the cleaning. The only thing I can
think of is that it's no longer an intermittent issue. That the primary
ports (or those two USB controller chips) have now weaken past the point
of intermittent. My most recent tests showed a consistency whereby,
using the primary USB ports there was always a Delayed Write Failed
error and there was no such error that occurred with the secondary
ports. I'm by no means saying this to establish a conclusion, it's just
an observation.

One other observation that might be worthy of note is that despite the
age of the P4C800-E Deluxe motherboard, the ASUS supplied USB plate was
put into service about a year ago. Those 4 ports that consist of the
secondary USB ports, are essentially brand new. While the primary and
the secondary USB ports are not of the same design, I noticed that it
requires a little bit more force to plug a USB cable into the secondary
ports.

[...]
That you really need to try something very basic like changing the USB
cable and see if you ever get any delayed write failures with a
different one. And it need to work fine for a week or more before you
can conclude thats where the problem lay.

As I mentioned, a different cable was tried. The USB cable that I used
belonged to a different brand and model of USB enclosure. I unplugged
the cable from the back of the other enclosure and plugged it into the
back of the Rosewill. Re-ran the test and received a Delayed Write
Failed error. However, at the time this different cable was connected to
one of the primary USB ports.
If you get a delayed write failure with a different USB cable you then
need to use the Vantec adapter for weeks or until a delayed write
failure shows up with that, to eliminate the Rosewell enclosure being
the problem.

Tomorrow I plan to take one of the hard drives out of the Rosewill and
connect it to the Vantec adapter. In the previous test I connected the
Vantec adapter to the monitor's hub, which in turn was connected to one
of the secondary USB ports. This resulted in no Delayed Write Failed
errors. This time around, I'll connect the Vantec adapter directly to
one of the primary USB ports. I'll leave it like that for awhile and
report back.
You have already shown that you do get delayed write failures on both
what you call the primary and secondary USB ports.

I'm not so sure that swapping one cable at a time was the best way to
test all of the ports.
 
R

Rod Speed

Tom wrote
Rod Speed wrote
[My Typo Corrected]
You cant be sure of that, because the fault is clearly intermittent and went
away for quite a while just by cleaning the Rosewill and system case.
True, it did go away for awhile after the cleaning.

And thats the evidence that it is an intermittent fault, because
cleaning cant make a delayed write fault go away.
The only thing I can think of is that it's no longer an intermittent issue.

Intermittent faults dont work like that.
That the primary ports (or those two USB controller chips)
have now weaken past the point of intermittent.

Chips dont fail like that. And that doesnt explain why you
have seen a delayed write failure on the secondary ports.
My most recent tests showed a consistency whereby, using the
primary USB ports there was always a Delayed Write Failed error
and there was no such error that occurred with the secondary ports.

Thats the problem with intermittent faults, they can fool you on the basics like that.
I'm by no means saying this to establish a conclusion, it's just an observation.
One other observation that might be worthy of note is that despite
the age of the P4C800-E Deluxe motherboard, the ASUS supplied
USB plate was put into service about a year ago. Those 4 ports
that consist of the secondary USB ports, are essentially brand new.

The physical port connectors dont age significantly.
While the primary and the secondary USB ports are not of the same design, I noticed
that it requires a little bit more force to plug a USB cable into the secondary ports.

Thats just basic clearance stuff. Its not significant.
As I mentioned, a different cable was tried. The USB cable
that I used belonged to a different brand and model of USB
enclosure. I unplugged the cable from the back of the other
enclosure and plugged it into the back of the Rosewill.
Re-ran the test and received a Delayed Write Failed error.

OK, it wasnt clear that thats what you had said.
However, at the time this different cable was
connected to one of the primary USB ports.

The USB port doesnt matter because you have
seen a delayed write failure on both types of ports.
Tomorrow I plan to take one of the hard drives out of the Rosewill
and connect it to the Vantec adapter. In the previous test I connected
the Vantec adapter to the monitor's hub, which in turn was connected
to one of the secondary USB ports. This resulted in no Delayed Write
Failed errors.

That does indicate the problem is with the Rosewell enclosure,
if it continues to work fine in that config for say a week or two.
This time around, I'll connect the Vantec adapter directly to one of the
primary USB ports. I'll leave it like that for awhile and report back.

It doesnt matter which port its connected to because you
have seen delayed write failures on both types of ports.

And you do need to do defrags frequently like say daily too.
I'm not so sure that swapping one cable at a time was the best way to test all of the ports.

There is no need to test all the ports since you have seen
delayed write failures on both primary and secondary ports.

So far it looks like the problem is with the Rosewell enclosure
and you need to run without it for a while and see if you ever
see any delayed write failures when it isnt being used.
 
M

Mike Tomlinson

Tom said:
Ack, that's a typo. I was attempting to convert 3 and 6 feet length
cables to meters in my head and inadvertently combined both without the
calculation.

No worries, but I still suggest you try a shorter, quality USB2 cable
(borrow one?) and try again.

Ignore anything Rod Speed has to say, he's the group idiot.
 
P

Paul

Tom said:
Anyway, at this point I can't get a Delayed Write Failed to happen as
long as the Rosewill sits on a secondary USB port. I've run a couple
other tests, such as running a MD5 check, which took hours to complete,
and there was no Delayed Write Failed. It's obvious that something has
changed with the primary USB ports and not for the better. What do you
think?

Thank you.

While your problem is likely interconnect, there is one other thing you
can play with.

In Device Manager, the "USB Root Hub" entries have an item in Power Management:

"Allow the computer to turn off this device to save power"

You can try removing the tick from the box, and see if things improve.

Paul
 
Ad

Advertisements

A

Arno

In comp.sys.ibm.pc.hardware.storage Mike Tomlinson said:
No worries, but I still suggest you try a shorter, quality USB2 cable
(borrow one?) and try again.

In theory the USB cable length should not matter that much
until you exceed 5 metres. If the cable is below spec or
if either side of the connection has damaged or below-spec
line drivers, it can matter. I have seen that in practice.

Grounding can also be an issue with longer cables for the
induced signal signal noise. With storage it should not matter
that much. Anyways, if the problems vanish with a 30cm cable,
then it is an electrical problem. If not, then you still do
not know.
Ignore anything Rod Speed has to say, he's the group idiot.

Well, sometimes he repeats useful answers others gave before
and they even match the question. Other than that, filtering
him out is a massive improvement to the signal/noise ratio.

Arno
 
M

Mike Tomlinson

Arno <[email protected]> said:
In theory the USB cable length should not matter that much
until you exceed 5 metres. If the cable is below spec or
if either side of the connection has damaged or below-spec
line drivers, it can matter. I have seen that in practice.

Me too, or people using cables supplied with USB1 devices with USB2
devices. USB2 cable is of a tighter specification. The answer is to
use the cable supplied with the device.

http://www.everythingusb.com/forums/showthread.php?threadid=809
filtering
him out is a massive improvement to the signal/noise ratio.

Certainly is. Never see anything he posts unless someone quotes him.
 
Ad

Advertisements

M

me

Your enclosure is bad as I had the same thing happen. Just buy a new
enclosure or see if Rosewill will swap it out. A relative had the same
problem and didn't know it and could have lost all data. I suggested he
buy a new enclosure and it worked perfect and didn't lose a byte. Good
Luck.



Tom said:
Hello.

I own a P4C800-E Deluxe motherboard, with Windows XP SP2 installed. I
realize it's an old motherboard, but I can't afford to purchase a new
computer. I'm hoping to get a few more years out of this one. :)

The issue has to do with a little over one year old Rosewill 'R2-JBOD'
external enclosure (self powered). The Rosewill is configured as JBOD,
with two Western Digital WD1002FBYS-02A6B0 1TB drives installed. The
Rosewill is connected directly to one of the motherboard USB ports.
Windows has the drives configured for quick removal and I always use the
'Safely Remove Hardware' to dismount the device. The Rosewill has worked
flawlessly for about a year in it's present configuration.

About one month ago, I went to defrag the Rosewill drives and several
minutes into the defrag I receive a pop-up that said there was a
'Delayed Write Failed', that data was "lost". I tried various things,
but in the end what I thought solved the issue was a thorough cleaning
of the inside of the computer case and the Rosewill enclosure. The issue
immediately stopped after the cleaning. At the time I used the drives
for approximately a couple more days without issue and the Rosewill was
then powered down.

One week ago I needed the Rosewill drives, and used them continuously
for four (4) days solid, until I needed to reboot the computer. After
rebooting, I decided to defrag the drives and I once again received a
'Delayed Write Failed'. I have been defragging the system and USB drives
with PerfectDisk for years without any issues.

Typically when there is a 'Delayed Write Failed' it says which folder is
"lost" and if I check that folder, all the files are gone. However, if I
reboot the computer, all the files are back and they are completely
intact. Which folder and which files are missing, change as I reboot the
computer. It's as if whatever is causing the issue, at that specific
point in time, is what's "lost". Also, most of the time I can hear the
Windows drive un-mount chime and immediately the mount chine -- like the
drive is momentarily dropping and coming back.

I've used CHKDSK when the 'Delayed Write Failed' has occurred and
especially when the files are supposedly missing and it reports that
there is nothing wrong with the drive. It seems whenever I use CHKDSK
nothing is ever found. Maybe I need better hard drive checking software,
what would you recommend?

The only other odd thing to happen recently was during one of the
reboots I received a BSOD. The error was 'UNMOUNTABLE_BOOT_VOLUME'
(Stop: 0x000000ED). According to Microsoft they say that this error
message occurs "When Volume on IDE Drive with Caching [is] Enabled".
They also say that this issue was fixed with Windows XP SP1, I have SP2.
I checked the boot drive with CHKDSK and CHKDSK again reports no errors.
And, it could be a coincidence, since in all these years it has only
happened twice, with a fair amount of time passing by and with several
boots in-between.

Some other things I've tried include:

1) Moving the USB cable for the Rosewill to one of the other seven
motherboard USB ports.

2) I've tried a different USB cable that's currently used for a
different brand of USB enclosure, which works fine with that drive.

3) I've removed the outer shell of the Rosewill to allow better air flow
and the metal hard drive mounting frame is only warm to the touch.

4) Rosewill fan is enabled.

5) I've cycled the USB connector into the Rosewill and motherboard ports
hoping to clear any possible oxidation. In addition, made sure that
the BIG and JBOD mode switch inside the Rosewill was firmly set to
JBOD by cycling the switch back and forth.

Other than these two issues, the computer is working fine. I don't know
if it's a motherboard (USB) issue or a enclosure (hard drive) issue.
Hence, posting to two Newsgroups, with followup set. Sorry for the
length of this post, I was trying to be as thorough as I could. If I
missed any important details, please let me know.

Thank you.
 
Ad

Advertisements


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