Now, you may think that there's enough disk copying software out there that I don't need to ask (from *ix dd to Partition Magic and Parted). But they all have one big problem
( Read more... )
Thanks, that looks interesting, and as long as the Linux version supports NTFS (which I think all do these days) that should be runnable from a CD 'live' disk image.
(It's just struck me that I could probably use gparted from an Ubuntu live disk to copy the WinXP system partition and create the data partition, then use the 'unstoppable' copier to copy the files on the data partition which is the one with the swap area and read fault.)
(As a result of a takeover and some other recent events, our work web access goes through a mess of filtering software called Websense. Livejournal is sort of accessible, but its stylesheets aren't, or something. I meant to follow up with more detail on ddrescue/dd_rescue at the time, but failed.)
This discussion is interesting to me, since I really should say something in Files that Last on how to recover a partially corrupted disk. I've marked this post in my LJ Memories. Thanks to those who've offered links.
Yes, particularly ddrescue can try to get data from several different backups of the file, and suggest having two or more copies of a CD so that hopefully they won't go bad in the same places. Both of them have a "try again later" ability (which for a hard drive might involve temperature change).
I have another drive on which this will be useful. It's only 20GB and hasn't been read for years, but even if I can only get a copy with the bad parts blank (but also known) I should be able to get the actual data off (it 'lost' the MBR, but since it's FAT partitions they should be fairly easy to find and recreate an MBR).
if you pass dd the `conv=noerror,sync` option it should skip over bad blocks (and write zeros to the corresponding position in the output). this is just academic, though, since as armb has noted, dd_rescue is the right tool to use in this case. (http://wiki.linuxquestions.org/wiki/Dd_rescue)
Hmm, as I recall there wasn't a noerror option back then (it was on kernel 2.2.19pre17 I tried it last, I think). Either that or it didn't work, still bombed on error. But thanks, that is a useful thing to know. So is dd_rescue, I wonder if that's in distros these days? [Checks] Hmm, it is in CentOS 5.4 but /isn't/ in Ubuntu 10.4LTS (last year's Long Term Service version, which is the one I use for rescue purposes). Bad...
http://dimg-tools.sourceforge.net/ might do it and there are, of course, load of tools that might be useful but don't give enought detail to tell at sourceforge
Comments 14
http://www.roadkil.net/program.php?ProgramID=29
as a starting point..
Reply
(It's just struck me that I could probably use gparted from an Ubuntu live disk to copy the WinXP system partition and create the data partition, then use the 'unstoppable' copier to copy the files on the data partition which is the one with the swap area and read fault.)
Reply
Reply
(The comment has been removed)
http://www.garloff.de/kurt/linux/ddrescue/
http://www.forensicswiki.org/wiki/Ddrescue
http://www.forensicswiki.org/wiki/Dd_rescue
For added confusion, at least one distribution has dd_rescue in a package called ddrescue, and ddrescue in a package called gddrescue.
Reply
Reply
Reply
I have another drive on which this will be useful. It's only 20GB and hasn't been read for years, but even if I can only get a copy with the bad parts blank (but also known) I should be able to get the actual data off (it 'lost' the MBR, but since it's FAT partitions they should be fairly easy to find and recreate an MBR).
Reply
Reply
Reply
Reply
Leave a comment