After shrinking my array, some of my user shares (not all) appear as empty.  

 

My steps included:

1. Moving the files to a different disk in the array using the unBALANCE plugin.

2. Removing the disk from the shares globally by deselecting in the include drop down. However I also excluded the disk globally, which based on the error I wasn't supposed to do.

3. Creating the new config.

4. Rebuilding parity with out the drive selected.

 

After reviewing my steps I see I should have deselected the disk for each share individually as well.

 

Is there, or what is, the process for repairing the user shares?

 

I have attached my diagnostics file.

 

Thanks, 

 

Willie

tower-diagnostics-20171110-0707.zip

Shares are just the top folders on each disk.

 

You can't see some data because you have filesystem corruption on disk2 (md2):

 

https://wiki.lime-technology.com/Check_Disk_Filesystems#Drives_formatted_with_ReiserFS_using_unRAID_v5_or_later

 

You should also consider converting all reiser disks to xfs.

 

 

 

 

New config does not affect user share settings. If you have globally excluded disks, and then assigned new disks to the same slots in the new config, you may need to un-exclude the disks.

 

Just now, johnnie.black said:

You should also consider converting all reiser disks to xfs.

 

Strongly agree!

 

Thanks for your input.  I

 

I ended up running the Reiserfsck --rebuild-tree command.  It has worked well enough to resurrect my user shares with minimal data loss.  The drive  is still disabled and is being emulated.  If I understand the next step correctly, it would be to create a New Config, leaving 'parity is valid' checked in order to re-enable the drive; after that I would re-run a parity check.  Is that accurate?

 

Thanks

 

 

If you ran the reiser_fsck against the emulated drive then those are NOT the correct steps.   In such a case you will need to rebuild the physical drive from the emulated one.   To use the same physical drive the sequence of steps is:

• stop the array and Unassigned the current drive marked as failed.
• start the array with no device assigned.    You should still see the data as the drive is being emulated, but UnRAID will ‘forget’ the serial number of the drive assigned to that slot.
• stop the array and reassign the drive.     You will be told that starting the array will cause the disk to be rebuilt.
• start the array to rebuild the physical drive to match the emulated one.

Thanks to all.  Array is up and running with very little data loss.

 

The excluded drive is still in my server residing as un-assigned.  Is there a plugin I can use to mount it to view with independent of the array?

 

Thanks, 

Unassigned Devices plugin.

6 minutes ago, johnnie.black said:

Unassigned Devices plugin.

 

But re-reading this thread why do you still have a disable disk? And when did it got disable, all disks were OK on the diags.

It is not disabled, it is un-assigned and will be physically removed shortly to make room for a second parity drive.  I would like to check if any of the data on it was data that was lost during the subsequent drive failure.