MSFN Forums > Is my Hard Drive near his death?

Full Version: Is my Hard Drive near his death?

MSFN Forums > Software, Hardware, Media and Games Central > Hardware Hangout

Fredledingue

Dec 22 2007, 03:42 PM

"D" is my second hard drive which has problem recently exposed.
When DMA is enabled on this drive, the computer freezes sometimes, when I try to access files on it.
Running scandisk under Windows froze the machine as well.
Running scandisk under DOS showed "Scandisk encountered a data error while reading FAT on Drive D. This prevents ScanDisk from fixing this drive. [Exit]"

Then I disabled DMA, and scandisk seemed to work under Windows.
It took about one hour to scan the first 70,000 clusters, then 2 hours to do 500,000 clusters, out of 1,200,000.
I stopped the process at about half the cluster scan because it was too late in the evening. I will re run the check tomorrow fully.

Here is the ScanDisk.log:

QUOTE ("ScanDisk")

*******************

Microsoft ScanDisk for Windows

NOTE: If you use an MS-DOS program to view this file, some of the characters
may appear incorrectly. Use a Windows program such as Notepad instead.

Log file generated at 20.12 on 12/22/2007.

ScanDisk used the following options:
Thorough test
Check dates and times
Automatically fix errors

Drive (D:) contained the following errors:

ScanDisk could no longer write to or read from sector 1376 of this drive's file allocation table.
Resolution: Repair the error
Results: Error was not corrected.

ScanDisk could not properly read from or write to cluster 3902,
which contains some or all of D:\temp\BATCH0.AUP.
Resolution: Repair the error
Results: Error was partially corrected.

ScanDisk could not properly read from or write to cluster 156434,
which contains some or all of D:\temp\BATCH5.AUP.
Resolution: Repair the error
Results: Error was partially corrected.

ScanDisk could not properly read from or write to cluster 157267,
which contains some or all of D:\temp\BATCH10.AUP.
Resolution: Repair the error
Results: Error was partially corrected.

ScanDisk could not properly read from or write to cluster 188752,
which contains some or all of D:\temp\BATCH20.AUP.
Resolution: Repair the error
Results: Error was partially corrected.

ScanDisk could not properly read from or write to cluster 190631,
which contains some or all of D:\temp\BATCH22.AUP.
Resolution: Repair the error
Results: Error was partially corrected.

ScanDisk could not properly read from or write to cluster 192842,
which contains some or all of D:\SysRest\71209AAB.ARJ.
Resolution: Repair the error
Results: Error was partially corrected.

ScanDisk was canceled.

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

Can someone tell me what "Error was partially corrected " means?
And if I should buy a new Hard Disc ASAP, before the datas on it became totaly lost.
I still didn't try to read files on this drive by fear of crashing my PC.

Thanks for your help.

Kelsenellenelvian

Dec 22 2007, 03:50 PM

UNPLUG IT!! Don't use it untill you get a new drive to put the info on it.....

PassingBy

Dec 22 2007, 05:18 PM

Hmmm .....

Sounds like surface degradation or mechanical defect(s) to me ... Probably more prudent to not run another scandisk at this point. If the scan stutters all over the drive surface, it is possible that the read/write section (maybe amplifier/coil/etc etc) is wearing out.

IMHO, you should attempt to copy your most important documents first before running a scandisk again. If the previous scandisk has gone through the first 10%-20% of your drive, it "probably" has fixed the FAT and DIR (YMMV). After having copied those wanted documents then you can attempt to scan the drive again.

If the drive is a slim maxtor, its going to RIP soon. If its a seagate, data recovery is possible but will burn a big hole. If its a WD, a surface erase can be of some help. If its a Samsung, its odd coz they're tough.

Rgds

Sysdll

Dec 22 2007, 05:34 PM

Yes first back up your important data and get another drive. I have drives that have shown similar scandisk logs after power outages. I’ve had some success in making them useful again by using the manufactures original startup utility and then Scandisk from the Windows ME boot disk followed by a format from whatever OS startup disk you’re using. All previous information is lost this way but it’s worth a try before tossing the drive.

Fredledingue

Dec 22 2007, 06:13 PM

Thanks to all, for your replies.

Sysdll
I'v got a power outage recently, yet it was still working properly after that. The file allocation table could have been affected, I don't know.
Do you think that repeated power outages could physicaly damage the drive? Or it's just bad for the datas?
The irony is that I'v bought a power accumulator (Mustek) supposed to provide power for computers during outages: It works when you pull the plug out of the wall, it doesn't when the power shut down from the network. Money in the wind...

Anyway...
This drive is already over 5 years old. I won't mind buying a new one.
Going to disconnect it now.

RJARRRPCGP

Dec 22 2007, 10:45 PM

QUOTE (PassingBy @ Dec 22 2007, 07:18 PM)

If the drive is a slim maxtor, its going to RIP soon.

I agree. They often start with inconsistent sector access.

herbalist

Dec 22 2007, 11:07 PM

The last time I saw error messages like that from Scandisk, the drive was on its last legs. The problem came on with very little warning. By the time I realized I had a problem, I'd already lost data. Managed to recover about 80% of the data by installing the drive as a slave and copying data a few folders at a time.

Have you tried Spinrite? I haven't used it but it's reputed to be good for data recovery and fixing errors.
Rick

Fredledingue

Dec 24 2007, 02:51 PM

Thank you for all your compassion for the dying hardware, especialy on this christmas eve.

I have copied about 9 Gb of the most useful datas to the other drive with one error here and there. About 99% of the datas have been saved out.
TotalCopy (aka Tcopy) has been of great help copiying the datas bypassing the errors without a crash.

Then I disabled the drive from bios. I'll disconnect it as soon as I figure out which one it is in the case. I suppose the one that is cold is the one to remove but the fan and the fresh air blowing to them is so efficient that there is no feelable difference of temperature.

Is it bad for hardware to be maintained too cold?

noguru

Dec 25 2007, 09:17 AM

Unless you are a Eskimo and live in a iglo, no. My WD is supposed to operate between 5°C and 55°C according to their own specs. Don't worry about that. It's not abnormal that a five year old drive gives up. Especially when used a lot. Sometimes it's starts with degrading performance and you can wait (and backup!) for the end to come. Just one power outage can be to much for a old drive.

Fredledingue

Dec 25 2007, 03:20 PM

Thanks, noguru!

My computer is in the cellar staircase, it's cold there but not under 0°C, and surely not under 5°C neither.
It's also a WD.

DonDamm

Dec 25 2007, 04:20 PM

Cold is generally good. If you value your data, then always make a back up before it crashes. The quickest way to do this for future reference is to make an image of the entire disk and save that somewhere else. Then you can always retrieve data if you really need it.

A power outage can indeed play havoc with the disk corruption and even the electrical control board on the drive itself. That may not be the case here, but you can't be sure without testing, so save your data first, then repair. Spin Rite is excellent for for a disk in your situation and wilol identify any bad sectors. It is like Scan Disk on steriods. It is slow, though, so set it to go overnight. If you can read data, then your partition table is probably okay. What happens to older disks is that the magnetic domains in some areas become a bit weaker and when the signals are right near the borderline, it becomes difficult for the drive to distinguish between a 1 or a 0. Those signals generate scan errors and the bad sectors are usually marked and avoided. As a disk ages, more and more of those can show up. ScanDisk and other utilities will mark these and then activate replacement sectors from an unused part of the disk. Most of this is transparent to the user until the errors become too numerous and the "spare" sectors too few.

Also, if the FAT is corrupted just a little, some of the file bits may have become lost or crosslinked. ScanDisk will find these and collect them into one file. Usually it is useless. If it's a five year-old disk I'm guessing it is no more than maybe 40GB. A brand spanking new 500GB Samsung will cost you no more than about US$80-90. I'd spring for a new disk and relegate this one to temp duty and not for important data.

Good luck with it!

Fredledingue

Dec 25 2007, 06:23 PM

Thanks DonDam,

There were indeed "crosslinked" files during previous scamdisk checks.

As soon as the store reopen, I'll get a new HD. It will be cool to have an extra 250Bg or something... and cope with the 138Gb limit.

RainyShadow

Dec 27 2007, 02:10 PM

Could you post the SMARTUDM report for this drive? There is a small possibility that the fault is in the cabling or the controller, not the drive. Already happened to me in the past.
Aside from this, try not to use the drive until you are ready to copy all data from it.

Fredledingue

Dec 28 2007, 04:07 PM

Here is the report.
What does it tell, Doctor?

CODE

SMARTUDM - HDD S.M.A.R.T. Viewer 2.00
                    Copyright © 2001-2003,  Sysinfo Lab
                    Copyright © 1997, Michael Radchenko
             www.sysinfolab.com   e-mail: support@sysinfolab.com

þ HDD 2 Model:          ST340016A
þ HDD 2 Size:           38166 Mb (37.27 Gb)
þ Location:             Primary Slave
þ Serial Number:        3HS2NY56
þ Controller Revision:  3.10
þ Buffer Size:          2048.0 kb
þ Compatibility:        ATA/ATAPI-5
þ PIO Mode Support:     4
þ SW DMA Mode Support:  None
þ MW DMA Mode Support:  2, Active: None
þ UDMA Mode Support:    5 (UltraDMA/100), Active: 5
þ Current AAM Value:    80h (80h recommended) - enabled
þ S.M.A.R.T.:           [û] disabled
þ SMART Self-test:      [ ]
þ SMART Error Logging:  [ ]

þ T.E.C. prediction monitoring started at: 28-12-07, 23:58:06
  Attribute                 ID Threshold Value Indicator  1/Month   T.E.C.
ÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ
* Raw Read Error Rate       1     34       65  þþþþþþ       0.0    Unknown
* Spin Up Time              3      0       70  þþþþþþþ      0.0    Unknown
  Start/Stop Count          4     20       98  þþþþþþþþþþ   0.0    Unknown
* Reallocated Sector Count  5     36      100  þþþþþþþþþþ   0.0    Unknown
* Seek Error Rate           7     30       87  þþþþþþþþþ    0.0    Unknown
  Power On Hours Count      9      0       82  þþþþþþþþ     0.0    Unknown
* Spin Retry Count          10    97      100  þþþþþþþþþþ   0.0    Unknown
  Drive Power Cycle Count   12    20       96  þþþþþþþþþþ   0.0    Unknown
  Drive Temperature         194    0       18  þþ           0.0    Unknown
  Hardware ECC recovered    195    0       65  þþþþþþ       0.0    Unknown
  Current Pending Sector    197    0      100  þþþþþþþþþþ   0.0    Unknown
  Uncorrectable Sector      198    0      100  þþþþþþþþþþ   0.0    Unknown
  UltraDMA CRC Error Rate   199    0      200  þþþþþþþþþþ   0.0    Unknown
  Write Error Rate          200    0      100  þþþþþþþþþþ   0.0    Unknown
  TA Counter Increased      202    0      100  þþþþþþþþþþ   0.0    Unknown
ÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ
NOTE: "*" means life-critical attribute
þ T.E.C. not detected.

  Attribute                 ID Threshold Value Worst Raw            Type
ÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ
* Raw Read Error Rate       1     34       65    58  0000092279A5h  PR ER
* Spin Up Time              3      0       70    70  000000000000h
  Start/Stop Count          4     20       98    98  000000000B95h  EC SP
* Reallocated Sector Count  5     36      100   100  000000000000h  EC SP
* Seek Error Rate           7     30       87    60  00001D3CDD39h  PR ER
  Power On Hours Count      9      0       82    82  000000003DF7h  EC SP
* Spin Retry Count          10    97      100   100  000000000000h  EC
  Drive Power Cycle Count   12    20       96    96  000000001247h  EC SP
  Drive Temperature         194    0       18    52  000000000012h  SP
  Hardware ECC recovered    195    0       65    58  0000092279A5h  ER EC
  Current Pending Sector    197    0      100   100  000000000021h  EC
  Uncorrectable Sector      198    0      100   100  000000000021h  EC
  UltraDMA CRC Error Rate   199    0      200   200  000000000000h  PR ER EC SP
  Write Error Rate          200    0      100   253  000000000000h
  TA Counter Increased      202    0      100   253  000000000000h  EC SP
ÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ
NOTE: "*" means life-critical attribute
Attribute types:
      PR - Performance-related         ER - Error rate
      EC - Events count                SP - Self-preserve

þ Reallocated Sectors:      0
þ Current Temperature:      18 C
þ Drive Power Cycle Count:  4679

PassingBy

Dec 28 2007, 08:13 PM

Hmmm .....

Seagate Barracuda IV ST340016A with 40GB capacity. The smart readings looks okay to me but as per your earlier description it does seems to have some difficulties with reading/writing. The reallocated numbers are small. There's a way to find out ...

Once you have copied out your data. Use any suitable utility to zero fill the entire drive then scan the drive again. If it seems to be smooth while scanning without much stuttering then its propably safe for normal use else its prudent to use it for secondary backup.

Rgds

Sysdll

Dec 29 2007, 12:35 PM

Repeated power outages can cause all kinds of errors. Power outages and spikes can also damage ram and bad ram will cause scandisk to report errors where there are none. Also when troubleshooting older computers (or any electronic equipment) it’s always a good idea to take it apart, clean it, and put it back together.

Fredledingue

Dec 29 2007, 02:12 PM

Thanks for your comments! And happy new year!

erpdude8

Jan 2 2008, 02:02 PM

Run Scandisk with the Thorough option in Win9x, or run Scandisk in DOS with the /surface option to check the hard drive for physical damage (aka. bad sectors). A LOT of bad sectors on the hard disk mean that the drive is about to be headed towards the afterlife and will need to be replaced.

If you live in an area where repeated power spikes occur I suggest you buy a UPS (uninterrupted power supply) unit. UPS units can supply power to your computer even in a black out, brown out or other power problems and prevent damage to your computer and other electrical appliances.

Happy New Year 2008 to everyone!

puntoMX

Jan 2 2008, 02:18 PM

QUOTE (Fredledingue @ Dec 28 2007, 04:07 PM)

What does it tell, Doctor?

My first impression is a dying motor and/or head-mechanism, or at least not powerful enough.

I would leave the drive as it is and use only your new 250GB, its way faster any way and I’m sure you won’t miss your 40GB.

RainyShadow

Jan 2 2008, 06:02 PM

SMART values are not too bad, but it seems to be disabled so these numbers might not be actual.

I'd suggest you get a copy of MHDD and use it to further diagnose the drive.
When you start the program and select the proper drive, first type "smart on" to enable SMART, then "smart aas" to enable attribute autosave. Then press F4 and perform a full scan with remmaping disabled . After it ends take a note of the bad sectors number (if any). If there are not much of these you could just scan again with remapping enabled, this will replace the bad sectors with spare sectors. If you do this, scan again to make sure that there are no more remaining bad sectors.

If there are a lot bad sectors ( more than 30-50) it's better to not remap them because you may run out of spare sectors and need them later. Better check for grouped bad sectors and repartition the drive leaving the bad areas between the partitions. You can later remap only single sectors left within partitions by performing a write operation on them (SMART remaps sectors only during writes).

You could also switch the DMA mode to UDMA/33 and see if that helps. Type "config" in MHDD and if there is a setting for the DMA mode change it. You have to power-cycle the drive for this to take effect.

puntoMX

Jan 2 2008, 08:46 PM

RainyShadow,

Thank you for the 2 links to the tools, very useful

This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.