Hi all

I need help recovering an Unraid 6.8.0 array after two disk failures.

My array has always used only the Parity 2 slot. Parity 1 has always been empty, and the array previously worked normally in this configuration.

Disk 7 originally used a 2 TB Western Digital drive. It became disabled, and its SMART report showed:

• 897 Current Pending Sectors

• 315 Offline Uncorrectable Sectors

Disk 9 originally used a 2 TB Seagate drive. It stopped identifying at the correct capacity and appeared as approximately 3.9 GB or 137 GB instead of 2 TB. Valid SMART data could not be read from it.

I used SystemRescue and GNU ddrescue to clone the failing Disk 7 to the new 8 TB drive.

The final ddrescue result was approximately:

• 2,000 GB rescued

• 99.99% rescued

• approximately 115,712 bytes unreadable

• approximately 113 bad areas remaining after retries

I bought a second new 8 TB drive to replace Disk 9.

The current intended assignments are:

• Parity 1: unassigned

• Parity 2: original 10 TB parity drive

• Disk 2 through Disk 6: original drives in their original slots

• Disk 7: ddrescue clone

• Disk 8: original drive in its original slot

• Disk 9: new blank replacement

I used Tools > New Config and selected Preserve Current Assignments: All. I then reassigned the Disk 7 clone and the new Disk 9 replacement.

Unraid still reports:

“Stopped. Invalid configuration.”

“Too many wrong and/or missing disks!”

The Start button remains greyed out.

I also tried 'mdcmd set invalidslot 0 9'

The command returned no output, but the Start button is still disabled.

I haven't started the array, formatted the replacement drive, or rebuilt parity.

My goal is to:

1. Accept the ddrescue clone as the valid Disk 7.

2. Preserve the existing Parity 2 drive as valid.

3. Rebuild only Disk 9 onto the new blank 8 TB replacement drive.

4. Add a new12TB Parity 1 drive. It's not currently assigned to the array. I don't intend to add or build the new Parity 1 drive until Disk 9 is rebuilt, unless the recommended recovery procedure specifically requires otherwise.

Can someone please confirm the correct recovery procedure for an array with only Parity 2 assigned, including whether invalidslot is appropriate and what the exact syntax should be?

I'll attach the latest diagnostics ZIP and screenshots of the current disk assignments and the disabled Start button.

Thanks!

Mark

plex-diagnostics-20260624-1851.zip

invalidslot not necessary these days.

1. Tools - New Config - Retain All - Apply

2. Assign disks as you specified

3. Very important! Check BOTH Parity Valid AND Maintenance Mode checkboxes then Start the array. This will accept all disks as assigned without altering any disks and without rebuilding Parity.

4. Stop the array, unassign disk to be rebuilt (disk9), Start the array in Normal (not Maintenance) Mode with nothing assigned as disk9. This will disable disk9 so it can be rebuilt.

Then post new diagnostics so we can see if emulated disk9 needs more work before rebuilding.

9 hours ago, mrichards said:

Unraid still reports:

“Stopped. Invalid configuration.”

“Too many wrong and/or missing disks!”

The Start button remains greyed out.

This implies you forgot to hit the Apply button during the New Config. After doing that all drives should end up with blue icons.

14 hours ago, trurl said:

invalidslot not necessary these days.

1. Tools - New Config - Retain All - Apply

2. Assign disks as you specified

3. Very important! Check BOTH Parity Valid AND Maintenance Mode checkboxes then Start the array. This will accept all disks as assigned without altering any disks and without rebuilding Parity.

4. Stop the array, unassign disk to be rebuilt (disk9), Start the array in Normal (not Maintenance) Mode with nothing assigned as disk9. This will disable disk9 so it can be rebuilt.

Then post new diagnostics so we can see if emulated disk9 needs more work before rebuilding.

Thanks so much, trurl. new diagnostics attached.

plex-diagnostics-20260625-1249.zip

Edited June 25Jun 25 by mrichards

There are two unassigned Seagate disks that appear to be failing; is this expected? Could also be a cable issue, but looks more like a disk issue, since there's not even a valid SMART report.

Emulated disk9 detects an xfs filesystem, but there are issues with ut; you can try check filesystem.

13 hours ago, itimpi said:

This implies you forgot to hit the Apply button during the New Config. After doing that all drives should end up with blue icons.

Thanks for responding, itimpi. I thought I had clicked Apply during the first New Config attempt, but based on the drive icons, it may not have completed or registered properly. I repeated the process using Retain All and Apply, and the drives then showed blue icons as expected.

Mark

5 hours ago, JorgeB said:

There are two unassigned Seagate disks that appear to be failing; is this expected? Could also be a cable issue, but looks more like a disk issue, since there's not even a valid SMART report.

Emulated disk9 detects an xfs filesystem, but there are issues with ut; you can try check filesystem.

Hi JorgeB

Yes, the two unassigned Seagate disks are expected. One is the original failed Disk 9, which is being detected at the wrong capacity and does not return a valid SMART report. The other is an older Seagate that is also being detected incorrectly. I am not planning to use either of them in the array.

I started the array in Maintenance Mode with Disk 9 still unassigned and ran: xfs_repair -nv /dev/md9

The result was:

Phase 1 - find and verify superblock...
bad primary superblock - bad CRC in superblock !!!

attempting to find secondary superblock...
found candidate secondary superblock...
verified secondary superblock...
would write modified primary superblock
Primary superblock would have been modified.
Cannot proceed further in no_modify mode.
Exiting now.

Nothing has been modified.

Thanks for your help with this.

Mark

Run without -n if it asks for it use -L

Post the output

1 hour ago, trurl said:

Run without -n if it asks for it use -L

Post the output

Are the attached files correct?

xfs_repair-disk9-verification.txt plex-diagnostics-20260625-2122.zip

Looks like you were still running with -n (nomodify).

1 hour ago, trurl said:

Run without -n

9 minutes ago, trurl said:

Looks like you were still running with -n (nomodify).

Argh! Sorry about that. I reran xfs_repair without -n and with -L. Output file and fresh diagnostics attached.

plex-diagnostics-20260625-2153.zip xfs_repair-disk9-repair-noN.txt

Restart in normal mode and see if the emulated disk mounts now and the contents look good.

6 hours ago, JorgeB said:

Restart in normal mode and

post new diagnostics

57 minutes ago, trurl said:

post new diagnostics

Hi JorgeB and trurl.

I restarted in normal mode. The emulated Disk 9 now mounts as XFS and shows 1.51 TB used with 6.49 TB free. I can browse the root and see the expected folders, though some folders are now empty. There is content in the lost+found folder.

Disk 7 still shows “Unmountable: Unsupported partition layout,” and Disk 8 still shows “Unmountable: No file system.” I have not formatted or rebuilt anything. New diagnostics attached.

plex-diagnostics-20260626-1012.zip

Edited June 26Jun 26 by mrichards

You'll have to do check filesystem on 7 and 8 then. And it does look like they have been unmountable since those initial diagnostics though you didn't mention it.

Did you not have a disk1?

Just now, trurl said:

Did you not have a disk1?

Did you ever have a disk1? If so, did you rebuild parity2 without disk1?

I guess if it is emulating disk9 well enough to make it mountable then parity2 must be valid enough.

When you eventually rebuild disk9 you will get those same lost+found. And you may end up with some lost+found when you repair disks 7 and 8.

Do you have any backups?

Linux 'file' command can try to identify what kind of data is in a particular lost+found file so you can try to open it in an appropriate application to see if you know what it is.

3 hours ago, trurl said:

Did you ever have a disk1? If so, did you rebuild parity2 without disk1?

I don't believe so, though I'm not sure why that would be.

2 hours ago, trurl said:

I guess if it is emulating disk9 well enough to make it mountable then parity2 must be valid enough.

When you eventually rebuild disk9 you will get those same lost+found. And you may end up with some lost+found when you repair disks 7 and 8.

Do you have any backups?

Linux 'file' command can try to identify what kind of data is in a particular lost+found file so you can try to open it in an appropriate application to see if you know what it is.

Is the next step to assign the new drive as Disk 9? Once that's done, I'll add the new 12TB drive as Parity 1.

re: backups.... I'd rather not answer that question. Please ask again after everything is running

Fortunately, the important files look like they're all there. I'll start backing things up ASAP.

3 hours ago, trurl said:

You'll have to do check filesystem on 7 and 8

3 hours ago, trurl said:

You'll have to do check filesystem on 7 and 8 then. And it does look like they have been unmountable since those initial diagnostics though you didn't mention it.

Did you not have a disk1?

Disk 7 is the ddrescue clone.

Yeah, I forgot mention Disk 8 because I was so focused on the other issues.

I ran check filesystem for both, though (attached). I really have to learn what these outputs mean.

I appreciate all of your help with this!

Mark

xfs_repair-disk8-check.txt xfs_repair-disk7-check.txt

12 hours ago, JorgeB said:

Restart in normal mode and

5 hours ago, trurl said:

post new diagnostics

3 hours ago, trurl said:

You'll have to do check filesystem on 7 and 8 then. And it does look like they have been unmountable since those initial diagnostics though you didn't mention it.

Did you not have a disk1?

Disk 7 is the ddrescue clone.

Yeah, I forgot mention Disk 8 because I was so focused on the other issues.

I ran check filesystem for both, though (attached)

1 minute ago, trurl said:

  12 hours ago, JorgeB said:

Restart in normal mode and

  5 hours ago, trurl said:

post new diagnostics

. I really have to learn what these outputs mean.

I appreciate all of your help with this!

Mark

2 minutes ago, trurl said:

  12 hours ago, JorgeB said:

Restart in normal mode and

  5 hours ago, trurl said:

post new diagnostics

plex-diagnostics-20260626-1509.zip

Sorry, I didn't notice that you didn't actually let it repair disks 7 and 8. You have to do it without -n

Had to use -L on Disk 8. Attached results for both just in case.

xfs_repair-disk7-repair.txt xfs_repair-disk8-repair.txt xfs_repair-disk8-repair-L.txt plex-diagnostics-20260626-1527.zip