Generate partial Disk Images

Question

I have a external HDD that

• produces random read failures (and it would after that show errors for all subsequent read access until I switch it off/on again)
• S.M.A.R.T information shows that the number of corrupted sectors is increasing (and I should replace the HDD with a new one)

Now my problem is that I want to generate one good RAW image .img file from this HDD and since my HDD is generating random "read sector failures" I want to find a program that is able to generate partial RAW disk images (that I can concatenate later to one complete good image).

I've tried so far:

1. Acronis True Image 2017 (commercial version)

   • This allows generating images (also split into e.g. 100 GB packages)
   • But it won't allow me to stop/continue after a read failure (and a powercycle of the HDD)
   • I can just retry (doesn't help) or ignore every "sector read failure" afterwards
   • Next time it would start from sector 0 again (so I never get more then the first 30-40% of the disk data)

2. HDD Raw Copy Tool

   • It advertises "low-level" copying, but the results were the same as above

Just from looking at command line tools like Linux's dd or some clones for Windows it's probably doable, but I would prefer something with a GUI on Windows which gives me some sort of visual control/overview of what I'm doing.

Alternatively I'm looking for a tool that does allow to "continue" after a failed disk-imaging at the position of failure (like with a download in my browser after I lost contact with the server).

Answer 1

After a long search and some trial and error e.g. with dd I ended up booting Linux for USB memory stick and using ddrescue command line tool.

For my case I was paricualry interested in an example from the manuals:

Example 5: While rescuing a partition in /dev/sda1 to the file hdimage, /dev/sda1 stops responding and begins returning read errors, causing ddrescue to mark the rest of the partition as non-scraped.

ddrescue -f -n /dev/sda1 hdimage mapfile     <-- /dev/sda1 fails here
  (restart /dev/sda or reboot computer)
ddrescue -f -n -A -i<pos> /dev/sda1 hdimage mapfile
  (if /dev/sda1 fails again, restart /dev/sda or reboot computer and
   then repeat the above command as many times as needed until it
   succeeds. <pos> is the position of the last block marked as
   non-scraped in the mapfile)
ddrescue -d -f -r3 /dev/sda1 hdimage mapfile

I've just replaced the - for files - unnecessary option -f (force overwrite) with -v (verbose) option.

My mapfile after serveral repeated second steps (and about 6 hours for 1TB later) looked like this:

# Rescue Logfile. Created by GNU ddrescue version 1.19
# Command line: ddrescue -v -n -A -i0xCCF9C10000 /dev/sda2 hdimage mapfile
# Start time:   ...
# Current time: ...
# Finished
# current_pos  current_status
0xE8D8BF0000     +
#      pos        size  status
0x00000000  0x59FAA45000  +
0x59FAA45000  0x0001B000  ?
0x59FAA60000  0x02B6A000  +
0x59FD5CA000  0x00016000  ?
0x59FD5E0000  0x5D600000  +
0x5A5ABE0000  0x00020000  ?
0x5A5AC00000  0x0CDB2000  +
0x5A679B2000  0x0001E000  ?
0x5A679D0000  0x2EDA9B000  +
0x5D5546B000  0x00015000  ?
0x5D55480000  0x3B5AB1000  +
0x610AF31000  0x6BE3F000  ?
0x6176D70000  0x8761E90000  +

And after the last call to retry and fill the gaps I ended up with just one none-recoverable sector and the following mapfile:

# Rescue Logfile. Created by GNU ddrescue version 1.19
# Command line: ddrescue -d -v -r3 /dev/sda2 hdimage mapfile
# Start time:   ...
# Current time: ...
# Finished
# current_pos  current_status
0x59FD5CA400     +
#      pos        size  status
0x00000000  0x59FD5CA400  +
0x59FD5CA400  0x00000200  -
0x59FD5CA600  0x8EDB635A00  +

References

• SuperUser: Recovering data from a hard drive which disappears from /dev upon reading data in a bad block
• Technibble: Guide to Using DDRescue to Recover Data