I began rescuing data from a 35 GB ext3 filesystem on a damaged hard drive by copying the corresponding device file with dd_rhelp, in order to obtain a disk image.

Unfortunatly, the drive I used didn't have enough space to hold both the log file, which permits dd_rhelp to manage dd_rescue and to resume its job, and the image. So, I had to stop dd_rhelp as it couldn't write anymore to the sane drive. The problem is that the disk image file is about 35 GB, but df reports only 3.7 GB used on the drive!

I got a bigger drive, and as I would like not to restart the recovery from the beginning, I copied the log and disk image files onto it. It seems both files were copied properly, but df still reports a wrong disk space usage on the new drive: 2.3 GB now, and of course the disk image file has still the same size (about 35 GB).

I'd like to know if I don't need to bother about what df says.

Thanks in advance,

Mathieu

How is inode usage?
df -i

Right now, I can't tell you, as I use another computer to access the hard drive.

I'll have a look at it. But, do you have any idea how it should look like in my case?

If you have 100% inodes used then it doesn't look good...
It looks just like a normal df screen but will tell you the number of inodes used instead of file space.

OK. For the "df -i" output, I had already tried on my laptop I'm using right now.

I guess for a 35 GB file (in fact, it's a little bit more), I should have 35*1024/4=8960 inodes used just for the file (I've formatted the 80 GB disk using one inode per 4 megabytes in ext2), plus a little bit more.

If the number of used inodes is correct, does this mean everything is all right?