Question bitrot hdds

I access my 2.5" external HDD one once a year: I connect it to my PC, check the S.M.A.R.T. data using CrystalDiskInfo or Kitfox, then disconnect it and put it back in my drawer.

Is it also necessary to perform a full surface scan once a year—using software like HD Tune or similar—to prevent “silent bit rot” (data corruption)?

I don’t believe scanning (reading) every sector of an HDD actually does anything in terms of “refreshing” the sectors.

The best defense against bitrot is redundancy, which can automatically repair it from a good copy or parity.

You have a single USB drive. Even if you’re using ZFS and a scrub finds corruption, you have no way to repair the bitrot. The best that you can do is know about the corruption, but that’s it.

…unless copies=2 is set. :winking_face_with_tongue:

My experience says otherwise:

Basically whence I activated ZFS scrubs twice a month, I’ve had ZERO bit rot on my NON-redundant media pool. In more than 6 years!

Prior to activating twice a month ZFS scrubs, I got a few bad blocks each year. Generally they were in video files, because they are both larger and I have a lot of them. So statistically more prone to bit rot.

Now back to the OPs question. Once a year full read is likely not often enough. Even for this “self correcting storage”. Perhaps every month or 2. Now if the drive used ZFS AND “copies=2”, then yes, less often scrubs would likely work.

I connect my 2.5" external HDD to my PC once a year to prevent bit rot (data corruption). Is it important and necessary to run a full surface scan (using some software)—a process that takes several hours? Does a full surface scan re-magnetize the files on the platter and prevent bit rot (corrupt data)?

if external hdd 2.5" copy files for hdd in 2023 for 2026 no execute full surface scan in years develop bitrot?

Have you ever posted anything to this forum that has anything to do with TrueNAS? No need to answer; you haven’t. Why are you here?

In what way do you believe that connecting a HDD to a PC once a year will prevent bit rot?

why not just backup more than 1 media - tape, blu-ray, floppy, zip drives, and my favorite punch cards. - and post PAR2 files online (just in case all of them go bitrot)

You forgot to link!

For a single drive that lives in a drawer, a surface scan tells you whether sectors can still be read, and unreadable ones show up in SMART anyway. What neither CrystalDiskInfo nor a surface scan can tell you is whether a file that reads back fine is still the same file you wrote in 2023. That’s the “silent” part.

The cheap way to catch it without ZFS is to save a list of checksums when you copy the files, then check against it each year. In PowerShell (assuming the drive is D:):slight_smile::

Get-ChildItem D:\ -Recurse -File | Get-FileHash -Algorithm SHA256 | Export-Csv $HOME\hashes-2026.csv -NoTypeInformation

Next year, run it again to a new file (same drive letter) and compare:

Compare-Object (Import-Csv $HOME\hashes-2026.csv) (Import-Csv $HOME\hashes-2027.csv) -Property Path,Hash

Anything listed has changed or gone missing. Keep the lists off the drive itself.

That only detects. As others said, repairing needs a second copy, so keep one on another drive or in the cloud, and when a file fails, restore it from the copy whose hash still matches your original list.

(Disclosure: I work on TimeBinder, which deals with file fingerprints. Nothing above needs it.)