3TB data drive turned RAW (win7)

[cccdfern]

3TB Drive turned to RAW (data, no boot files) I don't think I partitioned it beyond making it the proper size after buying since it came as a 1tb and required reformatting to become the proper size.

I was sorting things and tried to delete some pictures, the one I was deleting gave its preview, but all files after that did not. Interacting with any of these caused explorer to hang and prevented it from being ended through task manager. After some searching I ran checkdiskbut it hung halfway through and then after a reboot the drive became unmounted.

After that was I unable to get the computer to boot properly (hangs on windows splash screen) with it connected. Eventviewer was full of atapi & service control manager errors. Just in case I've only got one stick of ram in.

Found this through a stackexchange thread and ran it, selected analyse and then deep scan. Reading through others posts and the manual, it appears I may have made things worse.

Event Viewer:
https://drive.google.com/file/d/1-bFGmo ... sp=sharing

Log file:
https://drive.google.com/file/d/1r6XFDJ ... UXVqK/view

Currently mounted in top sata port of my machine and has been running testdisc for a few days. Had stopped at 22%. Computer Blue screened in the morning.

[cgrenier]

There are a lot of "I/O device error" messages in the log file.
You should clone the disk to a new empty one using ddrescue as described in https://www.cgsecurity.org/testdisk.pdf
Once it's done, remove the original disk and try to recover your data from the clone.
Good luck

[cccdfern]

RIght, managed to get fedora live going and install the ddrescue and the disk is actually showing up. The kicker is that I don't actually have another 3tb drive handy to copy to, there should only be a few hundred gigs on there, is there any way to copy that to a smaller one (1tb)? If not I'll order one.

[recuperation]

is there any way to copy that to a smaller one (1tb)? If not I'll order one.

No, because the goal is not to duplicate the content but to duplicate the volume/drive because you have to work on the full drive to reenable or to recover the content.
Upon succesful repair or recovery you are free to copy content to a smaller drive.

[cccdfern]

Okay, so I've gotten the new drive, formatted it and got back into fedora.

Code: Select all

Disk /dev/sda - 3000 GB / 2794 GiB - CHS 364801 255 63
     Partition			Start        End    Size in sectors
 1 P MS Reserved                   34     262177     262144 [Microsoft reserved partition]
 2 P MS Data                   264192 5860532223 5860268032 [Basic data partition] [backup3tb]
     NTFS, blocksize=4096

Disk /dev/sdb - 3000 GB / 2794 GiB - CHS 364801 255 63
     Partition			Start        End    Size in sectors
 1 P MS Reserved                   34     262177     262144 [Microsoft reserved partition]
 2 P MS Data                   264192 5860532223 5860268032 [Basic data partition] [Fox Box Media]
     NTFS, blocksize=4096

As per the manual, I just do ddrescue /dev/sdb /dev/sda sdb.log ?

[recuperation]

As per the manual, I just do ddrescue /dev/sdb /dev/sda sdb.log ?

As the syntax is
[...
ddrescue [options] infile outfile [logfile]
...]
you would overwrite the content of /dev/sda.

Your sda-device is not empty, though.

Used the -c parameter to improve the speed of copying. Try out higher figures like 1024,2048...

[cccdfern]

Well, it got to 99.99% in 4-5hrs, just ticking along now.

Code: Select all

[root@localhost-live ~]# ddrescue -f -c 2048 /dev/sdb /dev/sda sdb.log
GNU ddrescue 1.23
Press Ctrl-C to interrupt
     ipos:  852808 MB, non-trimmed:        0 B,  current rate:       0 B/s
     opos:  852808 MB, non-scraped:   34147 kB,  average rate:  35165 kB/s
non-tried:        0 B,  bad-sector:    8228 kB,    error rate:     170 B/s
  rescued:    3000 GB,   bad areas:      696,        run time: 23h 42m  7s
pct rescued:   99.99%, read errors:    16220,  remaining time:         n/a
                              time since last successful read:     25m 22s

[recuperation]

The huge amount of read errors (16220) is disclosing lots of dammage.
You are lucky to have a read speed of 35MB/s nevertheless.

[cccdfern]

Code: Select all

[root@localhost-live ~]# ddrescue -f -c 2048 /dev/sdb /dev/sda sdb.log
GNU ddrescue 1.23
Press Ctrl-C to interrupt
     ipos:    1052 GB, non-trimmed:        0 B,  current rate:       0 B/s
     opos:    1052 GB, non-scraped:        0 B,  average rate:  17357 kB/s
non-tried:        0 B,  bad-sector:   23095 kB,    error rate:     170 B/s
  rescued:    3000 GB,   bad areas:     2131,        run time:  2d  1m  6s
pct rescued:   99.99%, read errors:    45257,  remaining time:          0s
                              time since last successful read:         18s
Finished

And I can open the new drive and have started backing up to the nas.

Thank you both for the assistance.

[recuperation]

With 45000 read errors on the drive a lot of data and/or metadata may be lost.
Your duplicate should not have physical errors but logical errors everywhere where ddrescue failed to read the sectors from the source.
I would duplicate the drive again and the run chkdsk on the second duplicate.
The outcome will be free of logical errors in the file system.