Hi all,

This is on unraid 7.0.1

I'm looking for some support on how to replace two failing hard drives. Below is a snapshot of my array:

Disk 1 and disk 2 were beginning to show smart errors so I've ordered two replacement drives, they are however 16TB drives.

Disk 1:

Disk 2:

My thought process was to first replace the parity drive with a 16TB drive and let the parity rebuild, however during the rebuild I started getting the following syslog errors and 187 reported uncorrect started climbing on disk1:

Feb 25 12:23:50 AnneFrankNAS kernel: ata4.00: exception Emask 0x0 SAct 0x72011c7f SErr 0x0 action 0x0
Feb 25 12:23:50 AnneFrankNAS kernel: ata4.00: irq_stat 0x40000008
Feb 25 12:23:50 AnneFrankNAS kernel: ata4.00: failed command: READ FPDMA QUEUED
Feb 25 12:23:50 AnneFrankNAS kernel: ata4.00: cmd 60/c0:c8:30:96:24/02:00:6e:00:00/40 tag 25 ncq dma 360448 in
Feb 25 12:23:50 AnneFrankNAS kernel: ata4.00: res 43/40:c0:80:97:24/00:02:6e:00:00/00 Emask 0x408 (media error) <F>
Feb 25 12:23:50 AnneFrankNAS kernel: ata4.00: status: { DRDY SENSE ERR }
Feb 25 12:23:50 AnneFrankNAS kernel: ata4.00: error: { UNC }
Feb 25 12:23:50 AnneFrankNAS kernel: ata4.00: configured for UDMA/133
Feb 25 12:23:50 AnneFrankNAS kernel: ata4: EH complete

Feb 25 12:23:56 AnneFrankNAS kernel: ata4.00: exception Emask 0x0 SAct 0x7fe0001f SErr 0x0 action 0x0
Feb 25 12:23:56 AnneFrankNAS kernel: ata4.00: irq_stat 0x40000008
Feb 25 12:23:56 AnneFrankNAS kernel: ata4.00: failed command: READ FPDMA QUEUED
Feb 25 12:23:56 AnneFrankNAS kernel: ata4.00: cmd 60/c0:a8:30:96:24/02:00:6e:00:00/40 tag 21 ncq dma 360448 in
Feb 25 12:23:56 AnneFrankNAS kernel: ata4.00: res 43/40:c0:60:98:24/00:02:6e:00:00/00 Emask 0x408 (media error) <F>
Feb 25 12:23:56 AnneFrankNAS kernel: ata4.00: status: { DRDY SENSE ERR }
Feb 25 12:23:56 AnneFrankNAS kernel: ata4.00: error: { UNC }
Feb 25 12:23:56 AnneFrankNAS kernel: ata4.00: configured for UDMA/133
Feb 25 12:23:56 AnneFrankNAS kernel: ata4: EH complete

Feb 25 12:24:26 AnneFrankNAS kernel: ata4.00: exception Emask 0x0 SAct 0x7ff0060f SErr 0x0 action 0x0
Feb 25 12:24:26 AnneFrankNAS kernel: ata4.00: irq_stat 0x40000008
Feb 25 12:24:26 AnneFrankNAS kernel: ata4.00: failed command: READ FPDMA QUEUED
Feb 25 12:24:26 AnneFrankNAS kernel: ata4.00: cmd 60/28:a0:50:81:2e/02:00:6e:00:00/40 tag 20 ncq dma 282624 in
Feb 25 12:24:26 AnneFrankNAS kernel: ata4.00: res 43/40:28:40:82:2e/00:02:6e:00:00/00 Emask 0x408 (media error) <F>
Feb 25 12:24:26 AnneFrankNAS kernel: ata4.00: status: { DRDY SENSE ERR }
Feb 25 12:24:26 AnneFrankNAS kernel: ata4.00: error: { UNC }
Feb 25 12:24:26 AnneFrankNAS kernel: ata4.00: configured for UDMA/133
Feb 25 12:24:26 AnneFrankNAS kernel: ata4: EH complete

Feb 25 12:25:37 AnneFrankNAS kernel: ata4.00: exception Emask 0x0 SAct 0xff00001f SErr 0x0 action 0x0
Feb 25 12:25:37 AnneFrankNAS kernel: ata4.00: irq_stat 0x40000008
Feb 25 12:25:37 AnneFrankNAS kernel: ata4.00: failed command: READ FPDMA QUEUED
Feb 25 12:25:37 AnneFrankNAS kernel: ata4.00: cmd 60/28:c0:b0:17:46/02:00:6e:00:00/40 tag 24 ncq dma 282624 in
Feb 25 12:25:37 AnneFrankNAS kernel: ata4.00: res 43/40:28:c8:18:46/00:02:6e:00:00/00 Emask 0x408 (media error) <F>
Feb 25 12:25:37 AnneFrankNAS kernel: ata4.00: status: { DRDY SENSE ERR }
Feb 25 12:25:37 AnneFrankNAS kernel: ata4.00: error: { UNC }
Feb 25 12:25:37 AnneFrankNAS kernel: ata4.00: configured for UDMA/133
Feb 25 12:25:37 AnneFrankNAS kernel: ata4: EH complete

Feb 25 12:25:43 AnneFrankNAS kernel: ata4.00: exception Emask 0x0 SAct 0x3fcd8 SErr 0x0 action 0x0
Feb 25 12:25:43 AnneFrankNAS kernel: ata4.00: irq_stat 0x40000008
Feb 25 12:25:43 AnneFrankNAS kernel: ata4.00: failed command: READ FPDMA QUEUED
Feb 25 12:25:43 AnneFrankNAS kernel: ata4.00: cmd 60/e0:30:e0:28:46/01:00:6e:00:00/40 tag 6 ncq dma 245760 in
Feb 25 12:25:43 AnneFrankNAS kernel: ata4.00: res 43/40:e0:30:29:46/00:01:6e:00:00/00 Emask 0x408 (media error) <F>
Feb 25 12:25:43 AnneFrankNAS kernel: ata4.00: status: { DRDY SENSE ERR }
Feb 25 12:25:43 AnneFrankNAS kernel: ata4.00: error: { UNC }
Feb 25 12:25:43 AnneFrankNAS kernel: ata4.00: configured for UDMA/133
Feb 25 12:25:43 AnneFrankNAS kernel: ata4: EH complete

Feb 25 12:25:53 AnneFrankNAS kernel: ata4.00: exception Emask 0x0 SAct 0x3e00 SErr 0x0 action 0x0
Feb 25 12:25:53 AnneFrankNAS kernel: ata4.00: irq_stat 0x40000008
Feb 25 12:25:53 AnneFrankNAS kernel: ata4.00: failed command: READ FPDMA QUEUED
Feb 25 12:25:53 AnneFrankNAS kernel: ata4.00: cmd 60/40:48:a8:a4:46/05:00:6e:00:00/40 tag 9 ncq dma 688128 in
Feb 25 12:25:53 AnneFrankNAS kernel: ata4.00: res 43/40:40:08:a9:46/00:05:6e:00:00/00 Emask 0x408 (media error) <F>
Feb 25 12:25:53 AnneFrankNAS kernel: ata4.00: status: { DRDY SENSE ERR }
Feb 25 12:25:53 AnneFrankNAS kernel: ata4.00: error: { UNC }
Feb 25 12:25:53 AnneFrankNAS kernel: ata4.00: configured for UDMA/133
Feb 25 12:25:53 AnneFrankNAS kernel: ata4: EH complete

Feb 25 12:26:24 AnneFrankNAS kernel: ata4.00: exception Emask 0x0 SAct 0x7c03003f SErr 0x0 action 0x0
Feb 25 12:26:24 AnneFrankNAS kernel: ata4.00: irq_stat 0x40000008
Feb 25 12:26:24 AnneFrankNAS kernel: ata4.00: failed command: READ FPDMA QUEUED
Feb 25 12:26:24 AnneFrankNAS kernel: ata4.00: cmd 60/b0:80:18:07:54/02:00:6e:00:00/40 tag 16 ncq dma 352256 in
Feb 25 12:26:24 AnneFrankNAS kernel: ata4.00: res 43/40:b0:48:07:54/00:02:6e:00:00/00 Emask 0x408 (media error) <F>
Feb 25 12:26:24 AnneFrankNAS kernel: ata4.00: status: { DRDY SENSE ERR }
Feb 25 12:26:24 AnneFrankNAS kernel: ata4.00: error: { UNC }
Feb 25 12:26:24 AnneFrankNAS kernel: ata4.00: configured for UDMA/133
Feb 25 12:26:24 AnneFrankNAS kernel: ata4: EH complete

Feb 25 12:26:44 AnneFrankNAS kernel: ata4.00: exception Emask 0x0 SAct 0xff SErr 0x0 action 0x0
Feb 25 12:26:44 AnneFrankNAS kernel: ata4.00: irq_stat 0x40000008
Feb 25 12:26:44 AnneFrankNAS kernel: ata4.00: failed command: READ FPDMA QUEUED
Feb 25 12:26:44 AnneFrankNAS kernel: ata4.00: cmd 60/40:00:78:af:5c/05:00:6e:00:00/40 tag 0 ncq dma 688128 in
Feb 25 12:26:44 AnneFrankNAS kernel: ata4.00: res 43/40:40:a8:b1:5c/00:05:6e:00:00/00 Emask 0x408 (media error) <F>
Feb 25 12:26:44 AnneFrankNAS kernel: ata4.00: status: { DRDY SENSE ERR }
Feb 25 12:26:44 AnneFrankNAS kernel: ata4.00: error: { UNC }
Feb 25 12:26:44 AnneFrankNAS kernel: ata4.00: configured for UDMA/133
Feb 25 12:26:44 AnneFrankNAS kernel: ata4: EH complete

I realise now I probably should've swapped disk1 to the old parity dive when I assigned the 16TB as parity? I've cancelled the parity sync as I don't want to risk losing the data on disk1 (it has data we definitely don't want to lose).

Could someone please walk me through how best to proceed? When I stop the array and assign the old parity drive back as parity it obviously wants to parity sync, if I then stop the array again it auto populates the 16TB as parity and I'm unable to assign another drive for disk1, nor can I start the array with disk1 unassigned:

Hopefully I haven't borked it up already, thanks for any help!

Frank

Edited March 3Mar 3 by franksb89
Problem solved

1 hour ago, franksb89 said:

My thought process was to first replace the parity drive with a 16TB drive and let the parity rebuild

That was not a good idea; parity cannot be built correctly unless all drives are fine. Do you still have old parity intact?

9 minutes ago, JorgeB said:

That was not a good idea; parity cannot be built correctly unless all drives are fine. Do you still have old parity intact?

If I try to reassign the previous parity disk as parity it wants to start parity sync, so I assume no..

4 minutes ago, franksb89 said:

If I try to reassign the previous parity disk as parity it wants to start parity sync, so I assume no..

The question is whether you still have the old parity drive unchanged from when it was in the array? It is possible to force Unraid to accept the old parity drive without doing a parity sync. Have you written any new data to the array since trying to first replace the parity drive?

4 minutes ago, itimpi said:

The question is whether you still have the old parity drive unchanged from when it was in the array? It is possible to force Unraid to accept the old parity drive without doing a parity sync. Have you written any new data to the array since trying to first replace the parity drive?

Ah, no then I do think it's still intact. I replaced the parity drive with the new 16tb drive and let it run parity sync until it started showing errors. I then cancelled parity sync and tried to put back the old drive as parity. I have started the array with the old drive back in place but cancelled the parity sync as soon as it started running. No data was written to the array since.

New Config with all drives as they were originally, including the old parity drive. Before starting the array, check the Parity Valid box. That will prevent it from doing another parity sync.

Then come back, post your diagnostics, and ask us about the parity swap procedure.

18 minutes ago, trurl said:

New Config with all drives as they were originally, including the old parity drive. Before starting the array, check the Parity Valid box. That will prevent it from doing another parity sync.

Then come back, post your diagnostics, and ask us about the parity swap procedure.

Thanks everyone for the help so far!

I've done a new config with the original drives with the parity valid box checked. Seems like the array is back to normal now:

I've attached the diagnostics zip file:

frank-diagnostics-20260225-1611.zip

How do I go about starting the parity swap procedure?

Thanks again!
Frank

See the link: parity swap and use the old parity to replace disk1

Note that for the renbuild part, a filesystem check may be needed to correctly emulate disk 1 because parity may not be 100% in sync.

Hi all,

Thanks for the support. Parity has been copied and disk1 rebuild was succesful.

Next I need to replace disk2, before I replace it should I do a parity check?

It currently says parity is valid but the last parity check only took 6 seconds so I'm assuming this is from when I told it to accept parity with the new config?

The SMART report of disk2 isn't as bad as disk1 was, however this is how disk1 started before it got worse:

Thanks,

Frank

Post new diagnostics

Disk2 may be OK for a while; it's difficult to predict, but if the reallocated sectors continue to climb, then it should be replaced as soon as possible.

1 hour ago, trurl said:

Post new diagnostics

Uploaded!

frank-diagnostics-20260227-1758.zip

15 minutes ago, JorgeB said:

Disk2 may be OK for a while; it's difficult to predict, but if the reallocated sectors continue to climb, then it should be replaced as soon as possible.

Would you recommend doing a parity check before doing a rebuild on disk2?

Parity check shouldn't be necessary. Disk1 rebuild looks good and all drives mountable.

Are you planning to replace disk2 with a larger drive now that you have 16TB parity?

2 minutes ago, trurl said:

Parity check shouldn't be necessary.

Actually, since there was some chaos at the beginning, might be a good idea.

14 minutes ago, trurl said:

Parity check shouldn't be necessary. Disk1 rebuild looks good and all drives mountable.

Are you planning to replace disk2 with a larger drive now that you have 16TB parity?

Yea I have another 16tb drive to replace disk2.

11 minutes ago, trurl said:

Actually, since there was some chaos at the beginning, might be a good idea.

Alright thanks, I'll start a parity check now and replace disk2 tomorrow :)

Thanks again for the support everyone, both disks replaced and everything is back up!