Failed main board and 2 failed disks

[TChipperfield]

Long time reader first time poster, apologies if I do not list everything in the first post but I will try to capture everything.

 

I initially had a HP DL380 G8 server, all disks were sitting behind a disabled RAID controller, everything worked well.  I was away and something caused a failure, after some investigation it seemed I had two SAS drive failures and a mainboard failure.

 

I read a few articles at the time and (from memory) the best route was to move to new hardware as the replacement board was too expensive to replace.  When I moved to new hardware I had to create a new disk config as the drive serials appeared different due them previously being behind a RAID controller, I was unable to add the two failed SAS drives as they were completely dead plus the new hardware did not have a SAS controller.  I did/do have two parity drives and the data all appears to be in tact but I know I need to replace the drives as currently if any more fail then I know I am going to be screwed.

 

I have tried to pre-clear a new drive and add to the array but I get "Unmountable: Wrong or no file system" for the new disk.

 

Any advise or suggestions greatly appreciated. Diagnostic logs attached.

 

 

unraidhp-diagnostics-20230318-1601.zip

[JorgeB]

Log is full of spam, please reboot and post new diags after array start.

[TChipperfield]

Sorry JorgeB, as soon as I'm back home later today I'll reboot and get fresh logs. Thank you for looking.

[TChipperfield]

Sorry, bit later getting home.  I've rebooted and taken fresh logs. Hopefully these are not as cluttered. 

unraidhp-diagnostics-20230319-2056.zip

Edited March 19, 2023 by TChipperfield
Removed clutter

[JorgeB]

Check filesystem on disk1.

[TChipperfield]

Thanks JorgeB, I ran the check with the nv switch, result was "...Sorry, could not find valid secondary superblock."

 

I ran the above using the web GUI, should I try from SSH/Console? 

 

 

[JorgeB]

That is likely from the disk being previously on a RAID controller, but if no valid filesystem is found you'll need to re-format and restore the data from a backup.

[TChipperfield]

OK thank you, is there no way to rebuild the disk from the parity drives?

[JorgeB]

You can try, but parity would still need to be valid and the issue be partition related, which does not look like it.

[TChipperfield]

Really sorry to be asking lots of questions, I appreciate your time. How would I go about rebuilding the disk using parity?

[JorgeB]

Stop array, unassign that disk, start array, but likely the emulated disk will still be unmountable, and the array will be unprotected until the disk is rebuilt.

[TChipperfield]

I stopped the array and unassigned the disk, started the array and disk 1 now showing "Not installed". Stopped the array, added backed into to disk 1 slot, started array and now "Data-rebuild in progress"

In your experience, would you expect this to be successful?

[itimpi]

2 hours ago, TChipperfield said:

I stopped the array and unassigned the disk, started the array and disk 1 now showing "Not installed". Stopped the array, added backed into to disk 1 slot, started array and now "Data-rebuild in progress"

In your experience, would you expect this to be successful?

A rebuild will always restore exactly what was showing on the emulated drive before the rebuild was started.  

[TChipperfield]

Sorry if this is a stupid question but how do I see what is emulated on the disk before the rebuild?  Would the status of Disk 1 show something like "Emulated disk"?

[TChipperfield]

The disk rebuild completed, final message was that it completed successfully without errors but then we looking at the disk status it states "Unmountable: Wrong or no file system" but yet the disk rebuild was apparently successful. Is there anything that can be done about the wrong or no file system issue?  Thanks in advance

[JorgeB]

You can try checking filesystem on the disk again, but like mentioned you could have done that on the emulated disk to see if it was worth rebuilding.

[TChipperfield]

I'm so confused, I thought unraid should be able to recover from disk failures, have I done something wrong stopping the recovery?

[itimpi]

18 minutes ago, TChipperfield said:

I'm so confused, I thought unraid should be able to recover from disk failures, have I done something wrong stopping the recovery?

It should if you stayed on the same hardware.   You can normally even switch hardware as long as neither of them are using RAID controllers for the drives.   

[TChipperfield]

Understood, thank you for explaining. I thought I had done something or misunderstood the process.  I take regular backups so thinking I may start fresh the whole build and then copy the data back on.  Is there a guide on wiping/starting again from the same USB key?  I did have a search but nothing really along those lines.

 

Thank you again for the info.

[JorgeB]

Do you want to start completely fresh or keep some settings?

[TChipperfield]

Initially I was thinking start completely fresh but if there is an good way to keep the dockers then that would be great.  Thanks you

[JorgeB]

Copy the templates from /boot/config/plugins/dockerMan/templates-user, then point to the same appdata, docker image can be recreated.

[TChipperfield]

I don't fully understand the process above without context (apologies).  Is there a guide on just starting completely fresh?

[JorgeB]

To start completely fresh you just need to re-format the flash drive with a clean Unraid install, then copy your key, but make sure you make a backup first in case you need something later.