After attempting recovery after a drive / controller (PSU?) failure (see link below) I fired up Unraid and let it rebuild parity (disk 4 was marked bad). Now that the parity rebuild is complete it is still saying disk 4 is "Unmountable: No file system" (screenshot below). I assumed the parity rebuild would have taken care of this. Thoughts?

Please post diags.

Attached.

nas01-diagnostics-20190830-0750.zip

No valid reiser filesystem is being found on disk4, you can try again after changing fs to auto or if you're absolutely sure that's the filesystem it was you can try rebuilding the superblock:

https://wiki.unraid.net/Check_Disk_Filesystems#Drives_formatted_with_ReiserFS_using_unRAID_v5_or_later

 

 

18 minutes ago, johnnie.black said:

you can try again after changing fs to auto

Where do I change the FS to auto? 

 

I am positive it was reiser. Am I better flipping to auto or rebuilding the superblock?

Try auto first since no harm done, with the array stopped click on that disk, then change fs to auto.

Switched to auto - still saying unmountable. Switch back to reiser and rebuild superblock?

What about formatting and rebuilding (again) from parity? I guess parity has a corrupt FS for this disk ...

1 minute ago, autumnwalker said:

What about formatting and rebuilding (again) from parity? I guess parity has a corrupt FS for this disk ...

Parity has no idea of files, so you would simply rebuild the same contents again.    Do NOT format as that will write an empty file system to the disk and update parity accordingly guaranteeing the disk contents are lost.

Got it. So flip FS back to reiser and rebuild superblock?

I set the FS back to reiser and ran check, but got an error:

reiserfsck 3.6.27

Will read-only check consistency of the filesystem on /dev/md4
Will put log info to 'stdout'
Failed to open the device '/dev/md4': Unknown code er3k 127

 

You need to update to latest Unraid, IIRC that release has issues with reiserfsck.

4 minutes ago, johnnie.black said:

You need to update to latest Unraid, IIRC that release has issues with reiserfsck.

I have not updated due to the SQLLite DB corruption bug - I believe I am a use case for that. I did some searching on the error I am getting and see issues in 6.3.2 and reiserfsck 3.6.25.

 

The last update to reiserfsprogs that I see in change logs was 6.6.0 - are you sure 6.6.7 or 6.7.x have a more recent version?

8 minutes ago, autumnwalker said:

are you sure 6.6.7 or 6.7.x have a more recent version?

No, it's the same, try rebuilding the superblock directly, without checking first, start array in maintenance mode and:

 

reiserfsck --rebuild-sb /dev/md4

 

If it works follow the wiki instructions carefully, then will likely need --rebuild-tree.

 

 

root@nas01:~# reiserfsck --rebuild-sb /dev/md4
reiserfsck 3.6.27

Will check superblock and rebuild it if needed
Will put log info to 'stdout'

Do you want to run this program?[N/Yes] (note need to type Yes if you do):Yes
Failed to open the device '/dev/md4': Unknown code er3k 127

Just now, autumnwalker said:

Unknown code er3k 127

That's the same error the bad reiserfs tools gave, so likely another bug, you shouldn't be using reiserfs anyway, what happened to the old disk?

These drives (or their predecessors) were formatted when ReiserFS was the default option in Unraid. I've looked at the procedure to re-format and it's not all that appealing.

 

In the other thread (linked at the top) drives attached to my SAS controller were erroring. I think it was a PSU issue. The first time Unraid disabled Parity. The second time it disabled Disk 4. I put in a new PSU, started up the box, did the unassign drive, start array, stop array, reassign drive procedure to allow Unraid to rebuild the disk "failed" disk. When I reassigned the drive and fired up the array Unraid immediately showed it as unmountable, but also started to rebuild the array so I let it go. Now I'm here.

5 minutes ago, johnnie.black said:

what happened to the old disk?

Looking at the other thread it looks like you rebuilt on top, if the emulated disk was unmountable, you should have never rebuild on top of the old disk, if emulated disk is unmountable it will be the same after rebuilding.

Not many options now I'm afraid, you might ask for help on the reiserfs mailing list.

 

Reiserfs has been dead for a long time, all users are strongly recommended to convert to a different filesystem, for many years now.

So how do I get Unraid back up and running? Do I format the disk and forfeit whatever data was on disk 4? Assuming whatever is emulated right now is what I will be left with after format?

Just now, autumnwalker said:

Do I format the disk and forfeit whatever data was on disk 4?

You can do that, and other than the mailing list being able to help not many more options left.

1 minute ago, autumnwalker said:

Assuming whatever is emulated right now is what I will be left with after format?

Is my assumption correct?

1 minute ago, autumnwalker said:

Assuming whatever is emulated right now is what I will be left with after format?

There's nothing emulated at the moment all data on the other disks will remain, whatever data was on disk4 will be permanently lost.

I can see a couple of things missing (of any importance), but I do have a backup so I can restore them. About 1/4 of that disk was a TimeMachine backup which I can forfeit.

 

So when I powered the server back on the drive was disabled (red X) and showed unmountable. I did the un-assign / reassign procedure to have Unraid rebuild the disk from parity. This was my mistake? I was doing what I did in the last thread to have it rebuild the parity from disks, but that does not work in the other direction? 

 

What should I have done?

1 minute ago, autumnwalker said:

This was my mistake?

Yes.

 

1 minute ago, autumnwalker said:

I was doing what I did in the last thread to have it rebuild the parity from disks, but that does not work in the other direction? 

It would be fine if the emulated disk was mounting correctly, if it wasn't it was the wrong thing to do.