Re: [Bug-ddrescue] 250gig error... out of nowhere.

[Florian Sedivy]

Hi Niklas!

Am 05.11.2013 um 03:17 schrieb Niklas Swan:

I've tried a bunch of different commands, like attempting to run it backwards, direct mode (doesn't work on mac) but all of them either don't grab any data, or its just like "finished" and I'm like... no your not.

here's some info:

# Rescue Logfile. Created by GNU ddrescue version 1.17

# Command line: ddrescue -f -n -c 1 /dev/rdisk3 /dev/rdisk1 nsclone.log

# current_pos  current_status

0xAEAB74B400     +

GNU ddrescue 1.17

Press Ctrl-C to interrupt

Initial status (read from logfile)

rescued:   506054 MB,  errsize:    244 GB,  errors:   20248

Current status

rescued:   506054 MB,  errsize:    244 GB,  current rate:        0 B/s

   ipos:   750200 MB,   errors:   20248,    average rate:        0 B/s

   opos:   750200 MB,    time since last successful read:       1 s

Finished 

I tried starting it at 550gi, no dice... still finished.

so Questions:

is there any other DDrescue commands I can try to get the rest of the data off? 

(I refuse to believe 233gig was dead like that)

You said you tried different commands and I didn't see the log file, so I cannot be sure about it - but if you always included the option -n (--no-split) like you did in your example, ddrescue will only read non-tried areas. Since all of your drive has been tried before, you gave ddrescue nothing to do, so it finishes fast. 

I guess (and hope for you), the system only lost access to the drive, reporting all the remaining 234GB as one big read error. You could simply continue without the -n option, but then those 234GB would be read backwards 1 sector at a time (trimmed), which would be very slow. 

A better idea is to tell ddrescue that it should treat that area as non-tried. I can think of two ways to accomplish that: First, you could use -A (—try-again) together with an approximate -i 500G. This should reset the last 250200MB in the log file from "*" (non-trimmed) to "?" (non-tried). Or, even more precise, edit the log file yourself. The single large error should be the last line. Change the status character of that line from "*" to "?", save, and start ddrescue over. In both cases you can keep the -n option in for now if you want, but to finish the rescue you will eventually have to omit it. 

One final hint: -c 256 will give you optimum read speed from /dev/rdisk devices on OS X. 

And a word of caution: device numbers may change after unplugging and replugging drives. I always double-check those.

Good luck! 

Florian