0

I have a vmdk (VMware virtual [NTFS] formatted disk) beyond repair (It's not mine, I don't know exactly what happened).

I'm trying to get some files off of it using Plan B as outlined here: http://sanbarrow.com/sickbay.html#planb

From what I can tell, for my case this would entail

  1. Create a new new.vmdk with size a bit larger than the current one and format it NTFS
  2. Use ddrescue to copy the content of old.vmdk into new.vmdk
  3. Mount new.vmdk and try to recover stuff

My question is, how exactly would the ddrescue command look like? Assuming new.vmdk is sufficiently larger than old.vmdk and I want to start writing old.vmdk somewhere with an offset of 1024bytes, would this work?

ddrescue -o1024 old.vmdk new.vmdk

I have little margin for error, so I would like to confirm that I'm doing stuff right.

Bonus question: Does it matter what offset I use?

Thank you!

user9517
  • 114,104
  • 20
  • 206
  • 289
marius
  • 101
  • 2
  • 1
    Why do you have little margin for error? Can you not make a spare copy of your old.vmdk. How was the original vmdk damaged? Did the drive fail, was the filesystem corrupt, or did someone delete it or something? – Zoredache Mar 01 '11 at 20:14
  • 1
    What are the symptoms of "beyond repair?" – JakeRobinson Mar 02 '11 at 03:23
  • First principle of data recovery - Take an identical copy of the damaged media, and work from the copy. Never work directly on the damaged media. Hope this helps with the 'no margin for error' thing. – Chris Thorpe Mar 02 '11 at 08:04
  • I'm not sure this will help. If the filesystem on the old vmdk is shagged, then the copy you make will be equally shagged. – Tom O'Connor Mar 02 '11 at 09:02

1 Answers1

1

There are several vmdk mount utilities you may want to try one of those.

trent
  • 3,094
  • 18
  • 17