Hello, The other week I had a disk fail (disk 3), reported with a red cross in UNRAID and the disk was making an awful noise. I just replaced it and then started to rebuild the array from parity. Once the array had finished its rebuild from parity, another disk (disk 1) is now reporting as failed - I have done some extended smart checks and everything comes back without error.WDC_WD40EFRX-68N32N0_WD-WCC7K0JZJR4K-20250929-1943.txt

I attached the logs for the disk,

Please post the diagnostics.

As Requested ppnhost-diagnostics-20250930-0910.zip

It's not logged as a disk problem and SMART looks fine, recommend replacing the cables, and assuming the emulated disk is mounting and contents look correct, you can rebuild on top.

Ive checked the cables and they are all seated correctly, is there anyway i can clear the errors and recheck? how can i rebuild on top of the error is still present?

Post new diagnostics

ppnhost-diagnostics-20251001-0828.zip

Looks like you have connection problems with parity.

Have you checked all connections, SATA and power, both ends, including splitters?

Just rechecked everything, unplugged and pluged back in ppnhost-diagnostics-20251001-0828.zip

Those are the same diagnostics you posted earlier.

Sorry!ppnhost-diagnostics-20251001-1839.zip

Still seeing connection problems with parity

Oct  1 17:54:53 PPNHOST kernel: ata4.00: exception Emask 0x10 SAct 0x30000 SErr 0x400100 action 0x6 frozen
Oct  1 17:54:53 PPNHOST kernel: ata4.00: irq_stat 0x08000000, interface fatal error
Oct  1 17:54:53 PPNHOST kernel: ata4: SError: { UnrecovData Handshk }
Oct  1 17:54:53 PPNHOST kernel: ata4.00: failed command: WRITE FPDMA QUEUED
Oct  1 17:54:53 PPNHOST kernel: ata4.00: cmd 61/e0:80:88:60:e0/00:00:e8:00:00/40 tag 16 ncq dma 114688 out
Oct  1 17:54:53 PPNHOST kernel:         res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x10 (ATA bus error)
Oct  1 17:54:53 PPNHOST kernel: ata4.00: status: { DRDY }
Oct  1 17:54:53 PPNHOST kernel: ata4.00: failed command: WRITE FPDMA QUEUED
Oct  1 17:54:53 PPNHOST kernel: ata4.00: cmd 61/c8:88:70:61:e0/01:00:e8:00:00/40 tag 17 ncq dma 233472 out
Oct  1 17:54:53 PPNHOST kernel:         res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x10 (ATA bus error)
Oct  1 17:54:53 PPNHOST kernel: ata4.00: status: { DRDY }
Oct  1 17:54:53 PPNHOST kernel: ata4: hard resetting link

[4:0:0:0]    disk    ATA      WDC WD40EFRX-68N 0A82  /dev/sde   /dev/sg4 

    [parity] => Array
        (
            [idx] => 0
            [name] => parity
            [device] => sde

Have you tried replacing the cable?

Ive replaced the cable on Disk 1ppnhost-diagnostics-20251002-0924.zip

Now you're having connection problems with disk2

Oct  2 09:22:35 PPNHOST kernel: ata4: hard resetting link
[4:0:0:0]    disk    ATA      WDC WD40EFRX-68W 0A82  /dev/sde   /dev/sg4 
            [name] => disk2
            [device] => sde

Interesting, Its only disk one with the cross?

Yes, I know.

But before you can rebuild on top of disk 1, all other disks need to be reliably readable.

Make sure each cable has enough slack so it doesn't pull on the connection. Don't bundle data cables.

Im going to buy some more cables and replace what i currently have 1 by 1.

In the meantime i swapped one over Swapped the sata cable in disk 1 and the sata cable in parity. any changes in the logs.ppnhost-diagnostics-20251003-1526.zip

That looks OK so far. And emulated disk1 is mounted with plenty of contents.

On 9/30/2025 at 2:45 PM, p.wrangles said:

how can i rebuild on top

Stop array, unassign disk to be rebuilt, start array with nothing assigned to that slot.

Stop array, reassign disk to be rebuilt, start array to begin rebuild.

just started the rebuild, this looks bad

ppnhost-diagnostics-20251003-1904.zip

Disk 2 has completely disconnected. As mentioned, all disks must be readable to provide the parity calculated data for rebuilding a disk.

Maybe better wait until you get those new cables.

So what do i do now? just stop the rebuild?

yes

Am i at risk of loosing any data?