[Ted L. Nancy]

I am having trouble copying files off of one of my non-startup discs. The files are readable/accessible (I know this because they are old family movies that play just fine), but they won't copy. This is a major problem because... this is my only copy of these files-- I have no backup!

When I say the files won't copy, I cannot get them to copy to any other disc (I've tried several), or even to another folder on the same disc. I have tried Diskwarrior, including using its copy feature, but I get errors and essentially the files don't copy.

10.6.4 occasionally has given me copy error code -36. I did a bunch of google searches on this, and it seems this could be some type of error that 10.6.2+ cannot overcome due to sketchy disc formatting. Some people on other forums reported success after downgrading their OS to 10.6.0 or prior, but I'm not certain how to downgrade.

Also, only some of my files on this disc were affected. About half copied just fine. I was able to copy 1 file with Diskwarrior and 1 by compressing the file, copying that to another disc, and then decompressing. These "tricks" do not seem to be working with the remaining files.

Needless to say if this hard drive is going kaput, and I lose this stuff, then my wife will be devastated. Any help whatsoever is appreciated.

[Cold Warrior]

In 10.6.x I've also received the occasional Finder copy error. Force quit Finder and see if that helps. Usually works for me.

[Ted L. Nancy]

Cold,
For good measure, I just followed your advice, but to no avail. I think my issue is a bit more complicated than that.

Forgot to mention that Disk Utility found no issues with the disc. Going to try TechTool now...

[seanc]

Use SMART Reporter to get the SMART attributes of the disk, post them here.

Try putting the affected HDD into an external enclosure and trying it on another Mac.

[Ted L. Nancy]

2010-07-18 10:49:12.092 SMARTReporter[2551:207] Drive: 'ST31000340AS (Mac Pro HD | 9QJ1DQF6 | disk1)' Status: SMARTOK (S.M.A.R.T. condition not exceeded, drive OK)
2010-07-18 10:49:12.150 SMARTReporter[2551:207] Drive: 'ST31000340AS (Mac Pro Backup HD | 9QJ1F13C | disk3)' Status: SMARTOK (S.M.A.R.T. condition not exceeded, drive OK)
2010-07-18 10:49:12.201 SMARTReporter[2551:207] Drive: 'ST31000340AS (Movies-C HD | 3QJ0307D | disk2)' Status: SMARTOK (S.M.A.R.T. condition not exceeded, drive OK)
2010-07-18 10:49:12.755 SMARTReporter[2551:207] Drive: 'WDC WD10EACS-00ZJB0 (Movies-C Backup HD | WD-WCASJ1098636 | disk0)' Status: SMARTOK (S.M.A.R.T. condition not exceeded, drive OK)

(The drive at issue is "Movies-C HD.")

[seanc]

Not quite what I was looking for - I have had two of those WD10EACS drives crap out on me though...

Click the SMART reporter menu applet, go to preferences.
Under the Hard Disks tab, control click on the drive and select "Check SMART attributes".

Copy the section that looks like mine - my drive is a WD5000BEVT:

Code:
SMART Attributes Data Structure revision number: 16
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x002f   200   200   051    Pre-fail  Always       -       2
  3 Spin_Up_Time            0x0027   184   182   021    Pre-fail  Always       -       1766
  4 Start_Stop_Count        0x0032   099   099   000    Old_age   Always       -       1155
  5 Reallocated_Sector_Ct   0x0033   200   200   140    Pre-fail  Always       -       0
  7 Seek_Error_Rate         0x002e   100   253   000    Old_age   Always       -       0
  9 Power_On_Hours          0x0032   098   098   000    Old_age   Always       -       1593
 10 Spin_Retry_Count        0x0033   100   100   051    Pre-fail  Always       -       0
 11 Calibration_Retry_Count 0x0032   100   100   000    Old_age   Always       -       0
 12 Power_Cycle_Count       0x0032   099   099   000    Old_age   Always       -       1109
192 Power-Off_Retract_Count 0x0032   199   199   000    Old_age   Always       -       1033
193 Load_Cycle_Count        0x0032   135   135   000    Old_age   Always       -       197636
194 Temperature_Celsius     0x0022   112   098   000    Old_age   Always       -       35
196 Reallocated_Event_Count 0x0032   200   200   000    Old_age   Always       -       0
197 Current_Pending_Sector  0x0032   200   200   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0030   100   253   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x0032   200   200   000    Old_age   Always       -       0
200 Multi_Zone_Error_Rate   0x0009   100   253   051    Pre-fail  Offline      -       0

[Ted L. Nancy]

Sorry for screenshot image.. couldn't figure out ho to get that text to copy/paste.

[seanc]

Aye - your drive is toast!

Look at the re-allocated sector count! You'd think SMART would have tripped wouldn't you.
Pending/Offline sectors are at 24 which means it's managed to move some data to part of the disk that is in OK condition...

This is the same way my 2 x WD10EACS failed, at the same time, in a RAID 5 (yes I know they're not enterprise class drives blah blah blah).

Backup everything that you *can* - perhaps use the terminal to issue a cp -R command - copy recursively or perhaps rsync - besson3c is a good one for this, as it will get around the Finder throwing up when it hits a file it can't handle and should keep copying the other files.

Once you have done this, I'd suggest trying MHDD to scan the disk and attempt to fix the bad sectors - it's unlikely to be very successful but it's worth a shot - you may need a PC, never tried booting an MHDD CD on a Mac.

[Ted L. Nancy]

Frak!

Ok.. Data Rescue 3... would you recommend it?

[seanc]

I've only ever used it to recover a deleted file on the trial version, it worked... not sure how it would work for disk damage. A free alternative for recovering deleted files was Photorec (runs from the terminal), again, not sure of its usefulness on a failing disk.

I'd highly recommend MHDD, if the drive can be saved somewhat, MHDD will do it and may help you to access those specific files.

[OreoCookie]

Originally Posted by Ted L. Nancy 

Frak!

Ok.. Data Rescue 3... would you recommend it?

I haven't used version 3 yet, but version 2 has worked like a charm on three occasions. The only reason I effectively couldn't resurrect a backup drive is because I've tried Diskwarrior on it first. Then the files were still there, but all file names were lost. I highly, highly recommend Data Rescue.

Data Rescue needs an external harddrive, it just reads off broken volumes and thus it can't do any additional damage.

Even more importantly, invest money in an external harddrive and use Time Machine for backups. If you had backups now, you wouldn't even break a sweat, but just order a new harddrive, install it and copy your files from your backup.

[P]

The reason SMART hasn't triggered is that the "Value" in the Reallocated Sector Count is still 100, and so over the 36 threshold. I don't know how WD translates the raw value to their "value", but clearly something is off. Also, you have 24 sectors waiting to be reallocated, and an UDMA CRC error count that is also quite high.

Your quota of power-on hours compared to power cycle count is very high - implying that the drive has been going more or less 24/7 for a least 2 years. That's a lot. Consumer drives are not designed to take that. Try to get your data off using the terminal, then try dd for a block level copy (you need an emtpy second drive for that), then trash the drive.

[Ted L. Nancy]

Originally Posted by OreoCookie 

I highly, highly recommend Data Rescue.

Awesome. I'll give it a shot.

[Ted L. Nancy]

Originally Posted by P 

Try to get your data off using the terminal, then try dd for a block level copy (you need an emtpy second drive for that), then trash the drive.

Greek to me. How do I get the data off using the Terminal? What is "dd" and a "block level copy"?

[seanc]

Originally Posted by P 

The reason SMART hasn't triggered is that the "Value" in the Reallocated Sector Count is still 100, and so over the 36 threshold. I don't know how WD translates the raw value to their "value", but clearly something is off.

I come across many drives exhibiting file copy or access errors. These can have 2, hundreds or thousands of reallocated sectors and not be tripping SMART, I really don't understand what drive manufacturers are doing...

[P]

Originally Posted by Ted L. Nancy 

Greek to me. How do I get the data off using the Terminal? What is "dd" and a "block level copy"?

Yes, that went by a little fast, maybe.

In the Terminal, there are many tools to copy stuff. Easiest is off course plain old cp, for copy. IME, it's more tolerant than the Finder copy function.

dd is a tool to copy a disk completely, bypassing all filesystems etc. This can be useful to grab a copy of everything and then see what it looks like. It seems that most files have some bad data in them, but are mostly OK as they can be viewed. You will need to google the exact syntax for your uses, but basically you give it something like

dd if=/dev/disk0s2 conv=sync,noerror of=/dev/disk1s1

where the part after if is the location of the disk you copy from, and of is the disk you copy to. conv=sync,noerror means that the copy will continue even if it gets errors and pad the file to something of the corretc length. With video files, this usually means a bad fram or so, but it's viewable. To get the location of each volume, type "df" without the quotes.

The drive you copy to will be overwritten. Disconnect unnecessary drives before starting, and doublecheck that output location.

[Ted L. Nancy]

Ok thanks. That sounds like a sound procedure, but since I haven't typed stuff like that to control a computer since I was 8 years old using DOS, I'm gonna first see what Data Rescue can do for me.

Right now I've got both my problem drive and its intended backup drive disconnected from the computer and sitting in a cool, dry, safe place. I have a busy work week ahead, but when I get around to running Data Rescue, I'll post results. Many thanks.

[512ke]

Originally Posted by P 

snip Your quota of power-on hours compared to power cycle count is very high - implying that the drive has been going more or less 24/7 for a least 2 years. That's a lot. Consumer drives are not designed to take that.

Is this a recent thing? I have been running cpus and HDs 24/7 since 1986 with never a thought. Macs and PCs so scsi and ide. Are newer drives less able to stay on than in years past? What manufacturing processes make them less reliable?

[P]

Originally Posted by 512ke 

Is this a recent thing? I have been running cpus and HDs 24/7 since 1986 with never a thought. Macs and PCs so scsi and ide. Are newer drives less able to stay on than in years past? What manufacturing processes make them less reliable?

I think it's more economics than manufacturing processes. In 1985, most drives went to some sort of enterprise use. In 2010, most drives are for (what is labelled) consumer use, and the margins there are razor thin. Anything you can save on testing helps your bottom line.

There are some tradeoffs you can make as well. For instance, each sector on a drive has a large area before and after that isn't used to store data. The bigger that area, the more reliable the drive, but the less it can store. The ever decreasing physical size of drives has also made a difference. In 1985, drives were generally 3,5" full height drives. Today you can't get anything bigger than half height, and 2,5" drives are usually even lower. We used to call them slimline, but these days they're just referred to by their height in millimeters.

Bottom line is that it USUALLY works with consumer drives, but I generally don't take my chances. For something like a server, quality parts pay off - especially for things like PSU and HD.

[P]

Originally Posted by seanc 

I come across many drives exhibiting file copy or access errors. These can have 2, hundreds or thousands of reallocated sectors and not be tripping SMART, I really don't understand what drive manufacturers are doing...

I guess that's why they say that these conversion algorithms are proprietary...

Seriously, though, to still report 100 when you have 2000+ sectors reallocated is just stupid. Say that the alhorithm is linear. 2023 is not enough to trigger a switch to 99, so one step is at least 4047 (assuming a sane rounding algorithm). You need to lose 64 points to reach the threshold, so you need to lose some 261 000 sectors for SMART to trigger. That's 130 megs. How big is the freaking drive?

This is scary. I have a WD Black drive in this iMac, and if it uses the same algorithm as Ted's WD drive, it won't ever trigger SMART.

[seanc]

Well, it was Monday today and what did it bring? A 500GB Seagate 7200.11 with 1200ish bad sectors. Failing SMART? No. MHDD is running on it... I need to try and recover data.