Btrfs keeps reporting checksum failures in dmesg, including a rising corruption count. The system still seems stable, but the errors have come back across multiple drives. How can I narrow down the cause?
What should I check for recurring Btrfs checksum errors?
Answers
A checksum failure means Btrfs read data that did not match the stored checksum. That tells you something is wrong with the data read, but it does not identify which component caused it. A stable desktop does not mean the affected files are safe.
Back up anything you still need first, preferably to another device. Then run a Btrfs scrub on the mounted filesystem and check its results with btrfs scrub status. Scrub checks data and metadata, but it can only repair a bad copy when a usable redundant copy exists. Without one, you may need to restore affected files from a good backup.
Reply to this answer
Since this has happened with several drives, look beyond the SSD itself. RAM, the storage controller and power delivery are all worth investigating. A bad cable can be involved on systems that use one, though an M.2 drive normally connects directly to its slot.
Try a memory test and disable any XMP, EXPO or DOCP memory profile while troubleshooting. If errors continue at default settings, that helps narrow things down, but it does not prove the RAM is good.
Reply to this answer
Grab the SMART data too, even though the drive has been replaced. For an NVMe device, sudo smartctl -a /dev/nvme0n1 is a starting point. Check for media errors and other reported problems, and keep the full output rather than relying only on a pass or fail result.
Match the device name to the one in your logs first. A clean SMART report does not rule out RAM, controller or other hardware trouble.
Reply to this answer
Add your answer
Answers are reviewed before publication.
Veldaro+2n:
Scrub can also report existing damage rather than a new failure, so save its output along with the kernel messages for comparison.