advice for data recovery
Robin Laing
Robin.Laing at drdc-rddc.gc.ca
Thu Nov 13 18:33:29 UTC 2008
Dave Stevens wrote:
> Hello All,
>
> I am trying to recover .jpgs that were on a SATA drive that was formatted by
> mistake. No backup of course. Foremost has done a wonderful job of recovering
> several tens of thousands of files. Unfortunately many of them are either
> irrelevant (cached web fragments, etc.) or damaged. The most common type of
> damage is shown when trying to view them in Nautilus, when I get a message
> saying, "unsupported marker type."
>
> It seems about 30% - 40% of the files recovered are damaged in this way.
>
> They aren't my images and I am not able to gauge what is worthwhile or not but
> I would like to do some triage by only considering those of a certain minimum
> size (easy to do) and not damaged (no idea.)
>
> So does anyone know of a program I can use to only copy files that are not
> damaged? I can sort out the teenies, but don't see how to proceed after that.
>
> TIA.
>
> Dave
>
My sympathies. I did the same thing. I did a backup but forgot one sub
folder. Shame on me.
In the recovery, I ended up with 180Gigs of recovered files to go
through. Not pleasant.
I have not found a program that can check and confirm the accuracy of
the recovered files as they are all listed as images if I type in
file {file name}
What I have done is use "GQview" and look at the thumbnails.
My procedure is/was this.
With all the recovered images, I wrote a short bash script to sort them
into smaller sub-directories. This cut the number of images down to 10K
or 20K per directory.
Next I opened the directory in GQview. With thumbnails turned on, this
takes minutes, depending on the size of the folder.
I then sort the directory by file size. This helps to eliminate smaller
files from having to be viewed. I guess I could write a script to rm
these files anyway. See later.
Now with the thumbnails, I can quickly scan the directory for images. I
will use the mouse to drag and drop the file in the parent directory.
When scanning, the scrolling slows/stops for real images over corrupted
files.
I know my images from my camera are larger than 2M each but the recovery
software would pull any images it can and thus I sometimes have
recovered a smaller version of the image where the full image is gone.
I still have over 100Gig to go through.
I now backup my photos to DVD as soon as I take them off the stick.
One note about recovery software. As I did this at home, I cannot
remember what program I used. But what I did find out was it didn't
look at the headers for the images from my camera. I needed to change
the header info in a configuration file. I did a test on a memory stick
to ensure I was looking at my images.
--
Robin Laing
More information about the fedora-list
mailing list