I see no obvious hardware source for the problem, SATA Phy Event Counters are zeros except a few COMRESET events which is pretty minor. No ICRC or UDMA-CRC errors recorded. No reallocated or pending bad sectors.
I'd look at your historical system logs, messages or journalctl, and do case insensitive search for: err, unc, reset and see if anything along those lines has been reported that might explain it. If you decide to backup, reformat, restore, then first I suggest btrfs-image -c 9-t 4 /dev/sdX /mnt/<pathtoanothervolume+filename> You can keep it handy in case a dev asks for it or you can attach it to a kernel.org bug report. Chris Murphy-- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
