I have an array of 3x10TB disks (Parity, Disk1, and Disk2; both Disk1 and Disk2 have around 5TB of data each one, in total the array had a little more of 10TB in usage). Last week, Disk2 failed, but I was able to continue operating without any issues due to the parity protection.

On Friday, I received a new disk to replace Disk2 (the one that failed). However, when I tried to start the array with the  rebuild, it didnt  give me the option, because the system reported that the new disk was bigger than the parity disk (both are labeled as 10TB, but there may be slight differences in manufacturer specifications).

So I I followed the steps in the guide, starting from step 6 (as the earlier steps were already completed when the disk initially failed).

I progressed to step 14, which took around 15 hours, and then moved on to step 15, where I started the array and began the rebuild. At this point, the array became available, but I didn’t check whether all my data was intact. I also didn’t start any services, as I wanted to wait until the rebuild finished, which took another 15 hours.

After the rebuild finished, I started some of my services and noticed that I had lost some data—specifically al the newest files, that I think was previously stored on Disk2, which amounted to about 5TB. At this point, I returned to the Unraid dashboard to investigate. Disk2 (the newly rebuilt disk) showed no data usage or capacity. Instead, it displayed the message:

Unmountable: Unsupported or no file system after buld array

Since the data rebuild had completed some minutes ago (maybe a hour), I had no idea what this error meant or what was happening. I stopped the array and rebooted the server.
Upon restarting, Disk2 was marked as a "new device" and the array operations says start the array and rebuild again.

Now the disk is rebuilding once more, but I’ve already lost a significant amount of data.
 

My questions are:

- What did I do wrong? (I just follow the guide and use the WebUI)

- Did I lose the data because the parity-swap procedure or during the rebuild phase? Because when I was with the failed disk and the emulated disk2, the data existed

- Is there any way to recover the lost data?
 

Thanks for reading and helping me

Edited December 15, 20241 yr by RubenHC

Attach Diagnostics to your NEXT post in this thread. 

Have you tried doing a check filesystem on disk2 as the syslog suggests there is file system level corruption on that drive)?

I attach the disk2 xfs_repair status logs:

Edited October 26, 2025Oct 26 by Dosraid
Remove attached files

What's the next step? Run 'xfs_repair' without any parameters too fix de xfs filesystem?
Is the data already lost and cannot be recovered?

Diagnostics are after a reboot, so we can't see what happened with the parity swap, but for now, check filesystem on disk2, see the link and use the GUI, run it without -n

Thank you! After running the filesystem check (xfs_repair) without parameter and starting the array without the maintenance mode the data reappears.

@RubenHC

 

Check your lost+found share for anything repair couldn't figure out.