I've had a particular drive become disabled twice now.  After the first time, I tried to make sure the data and power cables were seated properly and booted back up.  The drive passed a SMART extended self-test, so I rebuilt the disk (without issue).

 

When it failed this time, I was moving files between 2 other array disks.  I'm basically sure nothing was writing to failed disk, but I set my md_write_method to "reconstruct write", so the disk was being read.

 

I plan to replace the disk with an unassigned drive already connected to the server on a different port, but I'd still like to know what caused this problem.  I'm attaching 2 diagnostics files:

 

20160907 - after 1st failure

20160909 - after 2nd failure

 

I'm also attaching 2 smart reports for the drive:

 

20160826 - well before first failure

20160907 - after 1st failure & after SMART extended

 

Can anyone tell me the most likely cause of this failure?  How should I go about confirming a point of failure?

sf-unraid-diagnostics-20160907-1255.zip

sf-unraid-diagnostics-20160909-1654.zip

SAMSUNG_HD204UI_S2H7J1BZ932669-diagnostics-20160826.txt

SAMSUNG_HD204UI_S2H7J1BZ932669-20160907-2124.txt

Before I dig through the diagnostics, what can you tell us about your hardware? How many disks? Controllers? If you have more than 5 disks, what is the exact model of your power supply?

Drives:

7 Data (including problem drive)

1 Parity

1 unassigned

1 SSD cache

 

Motherboard: Intel S2400SC2

 

All drives are connected to the motherboard, which has 6 SATA ports and 2 miniSAS connectors supporting 4 drives each.  I'm pretty sure the failed drive is on the 2nd miniSAS, which needed a storage activation key (RKSATA8) to function.

 

The power supply came with the system (Intel P4304SC2SFEN).  It's 550W. I'm attaching a picture of it.  I previously posted that I was concerned about how it would handle my drives, but that was more about a spike during spin-up.

Well, there are more syslog errors related to the drive, despite it being unassigned and unused (as far as I know):

 

Sep 10 18:27:02 SF-unRAID kernel: sas: Enter sas_scsi_recover_host busy: 1 failed: 1
Sep 10 18:27:02 SF-unRAID kernel: sas: trying to find task 0xffff88037eb6e400
Sep 10 18:27:02 SF-unRAID kernel: sas: sas_scsi_find_task: aborting task 0xffff88037eb6e400
Sep 10 18:27:02 SF-unRAID kernel: isci 0000:03:00.0: isci_task_abort_task: dev =           (null) (STP/SATA <NULL>), task = ffff88037eb6e400, old_request ==           (null)
Sep 10 18:27:02 SF-unRAID kernel: isci 0000:03:00.0: isci_task_abort_task: abort task not needed for ffff88037eb6e400
Sep 10 18:27:02 SF-unRAID kernel: isci 0000:03:00.0: isci_task_abort_task: Done; dev =           (null), task = ffff88037eb6e400 , old_request ==           (null)
Sep 10 18:27:02 SF-unRAID kernel: sas: sas_scsi_find_task: task 0xffff88037eb6e400 is done
Sep 10 18:27:02 SF-unRAID kernel: sas: sas_eh_handle_sas_errors: task 0xffff88037eb6e400 is done
Sep 10 18:27:02 SF-unRAID kernel: sas: ata12: end_device-2:1: cmd error handler
Sep 10 18:27:02 SF-unRAID kernel: sas: ata11: end_device-2:0: dev error handler
Sep 10 18:27:02 SF-unRAID kernel: sas: ata12: end_device-2:1: dev error handler
Sep 10 18:27:02 SF-unRAID kernel: ata12.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
Sep 10 18:27:02 SF-unRAID kernel: ata12.00: failed command: SMART
Sep 10 18:27:02 SF-unRAID kernel: ata12.00: cmd b0/da:00:00:4f:c2/00:00:00:00:00/00 tag 21
Sep 10 18:27:02 SF-unRAID kernel:         res 40/00:ff:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
Sep 10 18:27:02 SF-unRAID kernel: sas: ata13: end_device-2:2: dev error handler
Sep 10 18:27:02 SF-unRAID kernel: ata12.00: status: { DRDY }
Sep 10 18:27:02 SF-unRAID kernel: ata12: hard resetting link
Sep 10 18:27:02 SF-unRAID kernel: sas: ata14: end_device-2:3: dev error handler
Sep 10 18:27:02 SF-unRAID kernel: ata12.00: configured for UDMA/133
Sep 10 18:27:02 SF-unRAID kernel: ata12: EH complete
Sep 10 18:27:02 SF-unRAID kernel: sas: --- Exit sas_scsi_recover_host: busy: 0 failed: 1 tries: 1

 

The drive is still online, so I'm attaching a current SMART report, as well as full diagnostics just for completeness.  Any suggestions on what I should do next?

 

I successfully rebuilt the disk with another drive on a different port (but the same miniSAS motherboard connector).  I'm not sure yet if I want to use the array as normal or be more conservative for a while, since I'd really like to track down exactly what caused this problem.

sf-unraid-smart-20160910-2150.zip

sf-unraid-diagnostics-20160910-2150.zip