Everything as a bit rot monitor

[monke]

I've recently started experiencing error notifications whenever I connect a specific external drive:

image.png

Before proceeding with the chkdsk "repair" - which can lead to data loss, since it relies on flipping the blocks it deems broken - I would like to take a "blueprint" copy of the data. After the repair finishes, I'll take another copy and compare the two to see any potentially "rotten" data.

I could take a checksum of every file before and after the repair, but the drive in question is 18 TB, so it'll take a long time. I also don't want to stress a potentially compromised drive unnecessarily. Since Everything already maintains an existing, fairly complete record of the data on the drive (including the metadata, the MFT and the USN Journal), is it possible to have it detect any inconsistencies that occur? Or at least have it take two copies of the structure of the data to then compare them manually?

[void]

Everything can only really do this for filenames.

Export your files as an EFU file list from Everything -> File -> Export before the repair.
Then again after the repair.
Compare the two EFU file lists with a text diff tool.



Use something like voidhash to hash files before a repair.
The compare hashes after the repair.

[therube]

Doesn't (MS) Event Viewer also log detail of affected file (names, at least)?