Was adding disks to array, swapping out parity, physically moving devices in my box. In the process of all this starting and stopping of arrays and rebooting, I had a drive show up as missing and was listed under unassigned devices. The only option for this drive in unassigned devices is to format it, which I have not done as I have data on this drive.

 

I've swapped SATA cables to the drive as well as power to the drive. No change.

 

I see the below error:

Oct 28 14:23:01 Tower kernel: ata5.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
Oct 28 14:23:01 Tower kernel: ata5.00: irq_stat 0x40000001
Oct 28 14:23:01 Tower kernel: ata5.00: failed command: READ DMA EXT
Oct 28 14:23:01 Tower kernel: ata5.00: cmd 25/00:08:80:ff:ff/00:00:74:05:00/e0 tag 0 dma 4096 in
Oct 28 14:23:01 Tower kernel:         res 53/40:08:80:ff:ff/00:00:74:05:00/40 Emask 0x9 (media error)
Oct 28 14:23:01 Tower kernel: ata5.00: status: { DRDY SENSE ERR }
Oct 28 14:23:01 Tower kernel: ata5.00: error: { UNC }
Oct 28 14:23:01 Tower kernel: ata5.00: Read log 0x13 page 0x00 failed, Emask 0x1
Oct 28 14:23:01 Tower kernel: ata5.00: Read log 0x12 page 0x00 failed, Emask 0x1
Oct 28 14:23:01 Tower kernel: ata5.00: Read log 0x13 page 0x00 failed, Emask 0x1
Oct 28 14:23:01 Tower kernel: ata5.00: Read log 0x12 page 0x00 failed, Emask 0x1
Oct 28 14:23:01 Tower kernel: ata5.00: configured for UDMA/133
Oct 28 14:23:01 Tower kernel: sd 6:0:0:0: [sdf] tag#0 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=DRIVER_OK cmd_age=0s
Oct 28 14:23:01 Tower kernel: sd 6:0:0:0: [sdf] tag#0 Sense Key : 0x3 [current] 
Oct 28 14:23:01 Tower kernel: sd 6:0:0:0: [sdf] tag#0 ASC=0x11 ASCQ=0x4 
Oct 28 14:23:01 Tower kernel: sd 6:0:0:0: [sdf] tag#0 CDB: opcode=0x88 88 00 00 00 00 05 74 ff ff 80 00 00 00 08 00 00
Oct 28 14:23:01 Tower kernel: I/O error, dev sdf, sector 23437770624 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2
Oct 28 14:23:01 Tower kernel: ata5: EH complete

 

Any ideas on what's wrong? Thinking the disk may be fried but kind of surprised it just crapped out and hadn't thrown off any errors before this that I was aware of.

Please post the diagnostics.

diagnostics attached.

 

 

tower-diagnostics-20231029-1155.zip

That disk appears to be failing, you can run an extended SMART test to confirm, if it fails replace it.

47 minutes ago, JorgeB said:

That disk appears to be failing, you can run an extended SMART test to confirm, if it fails replace it.

 

Command "Execute SMART Extended self-test routine immediately in off-line mode" failed: scsi error medium or hardware error (serious)

 

Any way to recover the data on this drive since I can't start the array (was replacing a parity drive when this died, so no parity). Assuming drive is fried.

 

Additionally, if this drive is dead, can you confirm how I get the array back up without this disk, but maintaining the data on the remaining disks?

Is it tools -> new config -> preserve current assignments (all) that will put the array back up preserving the remaining data on the other disks?

Edited October 29, 20232 yr by privateer

6 hours ago, privateer said:

Any way to recover the data on this drive since I can't start the array (was replacing a parity drive when this died, so no parity). Assuming drive is fried.

Not without parity, if you still have the old parity maybe, if the disk was OK and parity is still mostly valid, you can also try ddrescue with the failing disk.

 

6 hours ago, privateer said:

Is it tools -> new config -> preserve current assignments (all) that will put the array back up preserving the remaining data on the other disks?

Yep.