At what point do you give up on ddrescue

[Haole Boy]

Aloha everyone. Yet another broken hard drive question....

Customer brought system to me with a "won't boot" issue. I booted it from an external drive, and SMART shows:

ID# ATTRIBUTE_NAME FLAGS VALUE WORST THRESH FAIL RAW_VALUE
1 Raw_Read_Error_Rate POSR-K 001 001 051 NOW 10523

All other SMART values are OK. And yes, I do know that SMART is not definitive in terms of drive health. This is a WD10SPZX-24Z10 drive (WD BLUE (EIDE)). Firmware version is 04.01A04, and I don't see a firmware update on the WD support site. The drive was in a Lenovo ideapad 330.

So I decided to pull the drive and run ddrescue. After 5 1/2 hours, the ddrescue info looks like this:

$ sudo ddrescue -d /dev/sde jennifer.img jennifer.logfile
GNU ddrescue 1.25
Press Ctrl-C to interrupt
ipos: 9492 MB, non-trimmed: 131072 B, current rate: 0 B/s
opos: 9492 MB, non-scraped: 0 B, average rate: 473 kB/s
non-tried: 990732 MB, bad-sector: 0 B, error rate: 0 B/s
rescued: 9472 MB, bad areas: 0, run time: 5h 33m 5s
pct rescued: 0.94%, read errors: 2, remaining time: 7d 17h 52m
time since last successful read: 0s
Copying non-tried blocks... Pass 1 (forwards)

Just before I captured this screen shot, the remaining time was 71 days, so it does look like some progress is being made, but it's really slow (less that 1% recovered after 5 1/2 hours). And there are only 2 read errors. I'm confused by this as I would expect lots of read errors for this very slow progress. The drive is in an internally mounted SATA dock.

My question is: am I causing more damage to the drive by continuing to run ddrescue? Or do I need to just be patient and let if finish - however long that might be?

Mahalo,

Harry Z

 

[Markverhyden]

What does the drive temp feel like? Internally mounted SATA dock? Meaning it's using SATA2SATA?

When I see huge jumps in estimated completion time it's usually a sign the drive is going to get worse relatively quickly. If you boot to a Linux distro can you browse the file system? If so I'd go that route to try to grab data.

 

[labtech]

Firmware version is 04.01A04, and I don't see a firmware update on the WD support site.
NEVER APPLY UPDATES on any HDD or SSD if it has failed & data recovery is needed. Updates do NOT solve issue for data recovery purposes, rather it prevents failure from a weird bug during software development (e.g. the HP SAS SSD suffering from the 40000 hours issue).
If there are good current backups in existence, then for failure prevention purposes, sure apply updates.

With your specific drive, something is wrong. More testing is necessary.
When scanning in MHDD for bad blocks, does it run smoothly OR does it run into bad blocks?

HDDGURU: MHDD

MHDD is the most popular freeware program for low-level IDE, Serial ATA and SCSI HDD diagnostics.

hddguru.com

 

[Diggs]

From what I understand (..and I stand to be corrected) ddrescue is just reading the drive and will try up to 10 times on a faulty sector before moving on and coming back later to read from the back side of the sector. When Windows tries to read the bad section of a drive it my hammer on it for 100s of trys so ddrescue is not any worse than what the drive has already seen. That said, I agree there is something weird with the drive if after 5 hours it only has 10 GB out of a TB recovered and showing no bad sectors.

BTW - @Haole Boy there is ddrescue GUI that puts a GUI on the front end of ddrescue if you want to get away from command line.

 

[lcoughey]

You didn't mention the brand or model of the hard drive. But, yes, it is likely that the underlying issues will only get worse as you proceed. Depending on the model and the underlying issues, it might be something that could be handled by RapidSpar.

I'd say that you should confirm with your client that the data is not worth sending off to a pro for possibly a few hundred bucks before you push it any further.

 

[Haole Boy]

I

What does the drive temp feel like? Internally mounted SATA dock? Meaning it's using SATA2SATA?

When I see huge jumps in estimated completion time it's usually a sign the drive is going to get worse relatively quickly. If you boot to a Linux distro can you browse the file system? If so I'd go that route to try to grab data.

Yes, I was booted to Linux, and did a manual copy of the user's Documents, Desktop, Photos, etc. folders. Much to my surprise, all copied without a problem. In fact, since the manual copy went so well, I rebooted into Windows and ran Fab's. Again, it worked without a problem. This drive is kinda strange...

And I'm using this 'enclosure' (link). Sorry I could not come up with the proper term last night

NEVER APPLY UPDATES on any HDD or SSD if it has failed & data recovery is needed.

With your specific drive, something is wrong. More testing is necessary.
When scanning in MHDD for bad blocks, does it run smoothly OR does it run into bad blocks?

HDDGURU: MHDD

MHDD is the most popular freeware program for low-level IDE, Serial ATA and SCSI HDD diagnostics.

hddguru.com

Thanks for the info about not updating drive firmware. Did not know that. Customer just wants to get their data, and I was able to run Fab's successfully , so I don't think I will pursue further diagnostics on this drive.

From what I understand (..and I stand to be corrected) ddrescue is just reading the drive and will try up to 10 times on a faulty sector before moving on and coming back later to read from the back side of the sector

In reading the online manual for ddrescue, if you don't specify the -r parameter, it tries each sector once. If you want multiple attempts you need to specify the -r parameter

I'd say that you should confirm with your client that the data is not worth sending off to a pro for possibly a few hundred bucks before you push it any further.

Yes, that is what I did, and (as usual) there was no interest in spending $400 or more for professional data recovery. My customer set is cheap.

Mahalo to all who replied. As always it is very much appreciated.

Harry Z

 

[frase]

My customer set is cheap.

..and there you have it. We can do this or that - soooo how long...is too long

I charge a minimal diagnostic fee, if client agrees all well and good. If not I just let them on their merry way to data oblivion.
One can tell before using DDrescue if the HDD is goin to have a chance or not, I check this first. Not only using SMART though multiple just to see diversity and for my own mind, completeness...ah the serenity.

I run a recovery not too intense incase of further reqs from Data recovery Specialists.





I run it for 24hours..if nothing well I advise client to a recovery specialist.

I set it all up for the client for recovery if going to a white room, advise costs and possibilities.

I have no time for ppl who want it all but pay minimal. I just say ok that's cool.

 

[Karlin High]

$ sudo ddrescue -d /dev/sde jennifer.img jennifer.logfile

You could try the -R option for "reverse," to start from the other end of the drive.

 

[Diggs]

You could try the -R option for "reverse," to start from the other end of the drive.

I know at least with GNU ddrescue it sweeps forwards and backwards twice as the default.