I have a drive pool using two Firecuda 2TB NVMe drives. I use this pool for my appdata (Docker), VM's, cached media (later saved to array), and system data. These errors don't seem to be generated during a parity check (parity is valid), I'm not even sure if the parity check only runs on the Array, or also on this pool.

The system is running ok, but I do experience weird glitches now and then. I'm more concerned that these errors keep increasing.

I'm not sure how to fix this and don't want to lose all my data. Does anyone know the process on how best to resolve these errors (see attached)?

Edited June 4Jun 4 by dougraid

Please post the diagnostics

Thank you @JorgeB here are the diagnostics.

centralperk-diagnostics-20260604-1449.zip

Btrfs is detecting data corruption on both devices. Start by running memtest, but I see you have a 13700K; it could also be the Intel 13/14 gen issue.

8 hours ago, JorgeB said:

Btrfs is detecting data corruption on both devices. Start by running memtest, but I see you have a 13700K; it could also be the Intel 13/14 gen issue.

I'm starting a memtest and will check the results once it has completed. I also have another intel processor (albeit a 14th Gen Intel) that I can swap out if you think it may be the current processor. However, let's say it was/is the processor causing the issues. If I swap that out, how do I fix the existing errors? Is it possible the drives themselves are bad, or is that not what you are seeing in the logs? Is there a text or fix I can run on the pool? Not sure where to go from here.

Unlikely that is a drive problem; once you believe the issue is resolved (note that it could also be an old issue, and the stats were never reset), scrub the pool, and post the results.

3 hours ago, JorgeB said:

Unlikely that is a drive problem; once you believe the issue is resolved (note that it could also be an old issue, and the stats were never reset), scrub the pool, and post the results.

Did a memtest, but not not seeing any errors. Did the scrub on the pool and these are the results:


UUID: 5dc4753d-9d19-4b7b-928f-d322f9130885

Scrub started: Fri Jun 5 14:34:52 2026

Status: finished Duration: 0:01:21

Total to scrub: 731.33GiB

Rate: 9.03GiB/s

Error summary: csum=8 Corrected: 0 Uncorrectable: 8 Unverified: 0

Still showing the Errors on the Pool. Is there a way to reset the errors? After switching out the processor, shat steps should I take from there?

11 hours ago, dougraid said:

Uncorrectable: 8

Look in the syslog for the list of correct file(s), delete/restore them from a backup, then run another scrub to confirm 0 errors and reset the stats, then keep monitoring the pool; if new errors show up, there's still a problem.