PhotoRec on Linux LVM2 disks - EXIF info is lost
Posted: 14 Aug 2018, 17:36
My Linux LVM2 Volume Group failed when 2 out of 4 disks returned an I/O error. File system is XFS, but it seems to be unrepairable, superblock cannot be found anymore.
I decided to use Photorec on each of the 4 standalone disks, the 2 good ones directly, the failing 2 by first running ddrescue. As many of the files on the LVM contain EXIF info, I was hoping to recover the file structure by using EXIF date info and original File ID.
PhotoRec has done a perfect job on the first disk, recovering many files. PhotoRec is now running on the second drive.
However, when I inspected the recovered files with ExifTool I noticed the important part of the original EXIF data was replaced by PhotoRec: the date info has been changed into the actual recovery date and the original FileID was replaced by the PhotoRec filename. This way I cannot restructure the files.
Would there be a possibility to run PhotoRec while preserving these EXIF tags (or did I miss something when reading through the documentation)?
And a more basic question: Is my approach of recovering data from a failed LVM the most appropriate one?
Thanks in advance for any help.
I decided to use Photorec on each of the 4 standalone disks, the 2 good ones directly, the failing 2 by first running ddrescue. As many of the files on the LVM contain EXIF info, I was hoping to recover the file structure by using EXIF date info and original File ID.
PhotoRec has done a perfect job on the first disk, recovering many files. PhotoRec is now running on the second drive.
However, when I inspected the recovered files with ExifTool I noticed the important part of the original EXIF data was replaced by PhotoRec: the date info has been changed into the actual recovery date and the original FileID was replaced by the PhotoRec filename. This way I cannot restructure the files.
Would there be a possibility to run PhotoRec while preserving these EXIF tags (or did I miss something when reading through the documentation)?
And a more basic question: Is my approach of recovering data from a failed LVM the most appropriate one?
Thanks in advance for any help.