[Help needed] Data recovery (media collection) from a mdadm array

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

Kristian

Active Member
Jun 1, 2013
347
84
28
Some of you may have seen my initial post:
https://forums.servethehome.com/index.php?threads/feedback-wanted-os-for-my-aio-homeserver-solved-recovery-of-a-crahed-xpenology-volume.6478/
So rubylaser was able to recover the recover my array with some effort (fsck, rebuilding the group descriptors, and the superblocks).

Of my original 21TB of data only 5TB seem to be recoverable instantly.
This includes the family photos, so I am ultimately happy.

For the rest I want to use something like Scapel or Photorec (tools I have never used before)

So now I would like to get back as much of the Data that could not be recovered as possible.

The content I am looking to restore is mainly recorded TV shows, our DVD and BlueRay collection, so nothing of big importance. (Leave aside the PITA to carry the 4 boxes (each weighting 30kg+ from the attic into the basement and let handbrake run for some months)

File formats will be mkv mostly.

But before I begin using Scalpel there are a few questions I would like to address:

1.) When I click right and open the properties of the array the following is shown:
Dropbox - Screen.jpg

Is this the normal behaviour?

Contents: 130.231 items, totalling 21,9 TB (some contents unreadable)
versus
5,2TB used and 34,4 TB free

I was under the impression that (some contents unreadable) was a problem of permissions.
If that is the case: Would this affect the used and free statistics?
It shouldn't is what I think.

So the file system knows the files are there at least?!
Does that make the operation with Photorec or Scalpel easier or harder?

I was guessing those tools are to recover „deleted“ files.
And the files seem to be all there


2.) When I browse the folders (I hadn’t had too much time yesterday and can continue in 12h the earliest) it seems that literally every file is present and most of them are shown with the correct size.
But some of them seem to be corrupted.

So I copy a 2gb video file of the array it is copied without errors, but can't be opened afterwards.
As you may imagine: I am not really in the mood to open all the files

(I can’t open them, or they open and a error message is shown that the stream could not be demuxed)
Is there a software that can verify if a video file is „okay“

See: I don’t want to copy the data of the array and notice later on, that 40% of the files are corrupted

So basically what I would like to do is:
Copy all files of the restored array that are okay and fully operational.
Begin the operation with scalpel to restore as much of the “unreadable contents” as possible.

Any thoughts? Input?

Best practices?
 

Attachments

Last edited: