Hi everyone,

I'm hoping someone can help me figure out a very strange array issue.

System

• Unraid 7.3.2

• 3 × 26TB WD drives

• Cache on NVMe

Current array

• Parity = WDC WUH722626ALE6L4 (sdc)

• Disk1 = WDC WUH722626ALE6L4 (sdd)

• Third drive = WD Gold WUH722626ALE6L4 (sdb)
  

The third drive is healthy. The disk previously contained data as an Unassigned Device, but has since been unmounted. No formatting has been performed.

Things already verified:

• SMART passes

• Linux detects the drive correctly

• No SATA/controller/I/O errors

• XFS filesystem mounts correctly

• Drive has been unmounted again

• Unassigned Devices no longer has the disk mounted

Current mdcmd status shows:

mdNumDisks=2
mdNumDisabled=1
mdNumInvalid=1

The array starts normally with parity and Disk1, but the third disk is never accepted into the array.

Kernel log contains:

md: 1 devices still in use.

but no hardware errors.

Could this be a corrupted array configuration (super.dat) or is there another known cause for mdNumInvalid=1 with an otherwise healthy disk?

Thanks!

PS: Already attempted:

• reboot

• stop/start array

• unmounted the disk from Unassigned Devices

• verified SMART

• verified filesystem

• disk detected correctly by Linux

• I have diagnostics available, but I'd prefer not to post them publicly unless they're really needed. If someone is willing to take a look, I'll happily provide them.

Please post the diagnostics

Thanks for your reply. I appreciate it. However, I'm not comfortable posting my diagnostics publicly. If that's essential to troubleshoot this issue, I'll leave it here. Thank you for your time.

Up to you, but can't really help without them.

Anonymize is an option when you create Diagnostics. And you can examine the diagnostics yourself, they are text. If you see anything in diagnostics you think needs to be anonymized that isn't, let us know.

While you get comfortable with that, post a screenshot of Main-Array Devices with the array started in Normal not Maintenance mode.

Thanks for the suggestion. I wasn't aware of the anonymize option when I created the diagnostics. I've now gone through them and sanitized the information I wasn't comfortable sharing publicly. I've attached the sanitized diagnostics below.

Here's the requested screenshot of Main → Array Devices with the array started in Normal mode.

bettycloud-diagnostics-20260803-1112_sanitized.zip

Thanks in advance for your time and efforts. I really appreciate it.

There appears to be a container continually restarting, spamming the log, and it is difficult to look for any other issues. Go to the Docker page and look at the container's uptimes; if you have one with just a few seconds, that is the problem, after resolving that (or sibsaling it), reboot and post new diags after trying to assign the disk.