unraid version 7.1.4.

My array:

parity: 6TB

disk 1: 4TB

disk 2: 6TB

disk 3: 2TB

disk 4: 2TB

disk 5: 2TB

Power outage. after booting, parity check started but couldn't complete because disk 2 was failing - tens of thousands of errors in SMART data but i didn't save the SMART data. i'm quite confident it died because to double check i reinserted the original disk 2, and unraid wouldn't even boot: the bios said "port 3: bad SMART data!". it boots just fine without.

i have a brand new 8TB HDD so i attempted the parity-copy operation by setting the original 6TB parity disk as disk 2, and setting the new 8TB disk as parity. the copy-parity operation keeps failing (after many hours of running) and i see these repeating errors in the logs:

Oct 19 17:30:57 unkar kernel: ata8.00: status: { DRDY ERR }
Oct 19 17:30:57 unkar kernel: ata8.00: error: { UNC }
Oct 19 17:30:57 unkar kernel: ata8.00: failed command: READ FPDMA QUEUED
Oct 19 17:30:57 unkar kernel: ata8.00: cmd 60/01:a8:ef:e5:a3/00:00:0e:02:00/40 tag 21 ncq dma 512 in
Oct 19 17:30:57 unkar kernel:         res 41/40:01:00:00:00/00:00:00:00:00/00 Emask 0x9 (media error)
Oct 19 17:30:57 unkar kernel: ata8.00: status: { DRDY ERR }
Oct 19 17:30:57 unkar kernel: ata8.00: error: { UNC }
Oct 19 17:30:57 unkar kernel: ata8.00: configured for UDMA/133
Oct 19 17:30:57 unkar kernel: sd 8:0:0:0: [sdh] tag#14 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=DRIVER_OK cmd_age=4s
Oct 19 17:30:57 unkar kernel: sd 8:0:0:0: [sdh] tag#14 Sense Key : 0x3 [current] 
Oct 19 17:30:57 unkar kernel: sd 8:0:0:0: [sdh] tag#14 ASC=0x11 ASCQ=0x4 
Oct 19 17:30:57 unkar kernel: sd 8:0:0:0: [sdh] tag#14 CDB: opcode=0x88 88 00 00 00 00 02 0e a3 e5 e8 00 00 00 01 00 00
Oct 19 17:30:57 unkar kernel: I/O error, dev sdh, sector 8835556840 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0

I changed the SATA cables and did two more attempts at copy-parity, which resulted in the same failure with the same errors.

Now, the SMART data for the 6TB parity has the following outstanding data:

reallocated sector count: 531

current pending sector: 1

this drive is probably also failing because the reallocated sector count was 231 before the power outage. However, when i run the array with the original config and the original disk 2 removed, and therefore emulated, everything is fine, and i am able to copy the disk 2 data manually from the array, without any unusual behavior whatsoever.

My secondary questions:

1. i'm fairly confident my parity drive is failing, but maybe someone has a better explanation about, or solution for, this behavior.

2. why is the hdd good enough to copy data from manually, but not enough to do a copy-parity operation, considering that both are read-only operations and should exert the same stress on the hdd? This sometimes-failing behavior is making me second-guess my diagnosis.

main question:

what is my best course of action to restore my array with the new 8TB HDD as the parity drive? Data is backed up and no issue, however, i'm looking for the least painful way to do this because everything is slow and restoring data will take a very long time. I understand that having lost two 6TB HDDs, my array will have a smaller capacity - this is fine.

I'm thinking of just making a new config without the original disk 2, with the new 8TB HDD as the parity drive. Eventually, i'll be able to get a new drive and bring the capacity back to what it was before.

i'm hoping someone has a better idea. diagnostics file attached.

unkar-diagnostics-20251020-1816.zip

2 hours ago, khindjal said:

the copy-parity operation keeps failing (after many hours of running) and i see these repeating errors in the logs:

That suggests parity is also failing, and SMART shows some issues; you can run an extended SMART test to confirm.

The results of the extended SMART test on the parity drive:

Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
# 1  Extended offline    Completed: read failure       20%     38171         245622248

So i'm guessing the parity drive has bad sectors, but coincidentally i have not run into them during normal operation.

Parity has also failed, so the parity copy will fail. One possible option would be to clone parity with ddrescue, then use the clone as the new parity, of course some data loss can still happen, depending on how many bad sectors there are and if they coincide with any data locations.

9 minutes ago, JorgeB said:

One possible option would be to clone parity with ddrescue, then use the clone as the new parity, of course some data loss can still happen, depending on how many bad sectors there are and if they coincide with any data locations.

Yes! this is exactly the kind of thing i was looking for! It fits my criteria of "least painful way". Thank you!

i'll get back with an update on how it went.

You will need to do a new config to use the cloned parity, and since there's a disabled disk, it needs a specific procedure, so please ask if you need help with that part.

i did ddrescue to clone the parity drive onto the new 8TB drive. Results of ddrescue:

GNU ddrescue 1.29.1
Press Ctrl-C to interrupt
Initial status (read from mapfile)
rescued: 6001 GB, tried: 4096 B, bad-sector: 4096 B, bad areas: 1

Current status
     ipos:        0 B, non-trimmed:        0 B,   current rate:       0 B/s
     opos:        0 B, non-scraped:        0 B,   average rate:       0 B/s
non-tried:        0 B,  bad-sector:     4096 B,     error rate:       0 B/s
  rescued:    6001 GB,   bad areas:          1,       run time:          0s
pct rescued:   99.99%, read errors:          0, remaining time:         n/a
                               time since last successful read:         n/a

Finished

So i guess there's just the one bad sector?

if you could help me with the procedure (keeping mind the disabled disk), that would be great!

29 minutes ago, khindjal said:

So i guess there's just the one bad sector?

Looks like it, there should be minimal data loss, if any. With some luck there's no data in the corresponding data disks.

To use that disk as the new parity:

-Tools -> New Config -> Retain current configuration: All -> Apply

-Check all assignments and assign any missing disk(s) if needed, you need to assign a temporary spare disk for the disabled disk, replacement disk should be same size or larger than the old one (but smaller than parity), assign the cloned parity in place of the old one.

-IMPORTANT - Check both "parity is already valid" and "maintenance mode" and start the array (note that the GUI will still show that data on parity disk(s) will be overwritten, this is normal as it doesn't account for the checkbox, but it won't be as long as it's checked)

-Stop array

-Unassign the disabled disk

-Start array (in normal mode now), and post new diagnostics, but you should now be in the same situation as before, but with a good parity.

Operation complete and everything is working as expected! thanks a lot for your help! this saved me many hours of moving data back and forth, I can accept one corrupt file.

New diagnostics file attached. thanks again!

unkar-diagnostics-20251022-1642.zip

Emulated disk2 is mounting, so you can now proceed with the parity swap, or just a normal rebuild if you have another spare disk that is not larger than parity.

i have no spare disks at the moment. Is it worth attempting a rebuild with the drive with the bad sector, or am i just asking for trouble?

It's difficult to say, most likely, the rebuild (full disk write) will reallocate that pending sector. The hard part is to guess if new one(s) will appear soon after, but it could still be better than having the array unprotected, e.g., if another disk fails in the meantime, at least disk 2 should be mostly recoverable.

20 hours ago, khindjal said:

Is it worth attempting a rebuild with the drive with the bad sector, or am i just asking for trouble?

You could try running the extended SMART test to see if that can complete without error (possibly after a preclear to act as a stress test and an attempt to force a reallocation of the bad sector). The short SMART test would unfortunately not be much use so anything will take some time.

i assigned the old parity drive to disk 2 in the new configuration. i was able to rebuild onto the drive with the bad sector just fine. in the SMART data, "current pending sector" has returned to 0. Read operations on the drive as part of the array are working just fine, and I've copied data onto the drive to test it and everything seems to write normally.

It appears that the bad sector has been reallocated successfully. It's working fine for now - and with a rebuilt parity i'm safe against it failing.