smartctl shows all disks passing. I'm not a SMART expert, so I can
post the results if you want to look at them yourself, but it seems
like a lot to post if you don't want it.
Check the following attributes:
- Reallocated_Sector_Ct
- Seek_Error_Rate
- Current_Pending_Sector
- Offline_Uncorrectable
- Raw_Read_Error_Rate
If any of these are increasing, its a sign of a dead drive. If
the drive is taking a long time to time out from a surface defect it could
be the cause.
- UDMA_CRC_Error_Count
That one is usually a sign of a bad cable.
Are the components of md1 (the unrelated array) on a different hardware
controller / wires?
Same controller, but I see the same results even if I unplug md1.
Have you tried replacing the cabling and the backplane?
I'm not sure what layout you have, but something I'd try is to remove the
drives for md1 and put the md127 drives in the place where the md1 were.
It would help rule out that either is the issue.
