Array down following reboot - Too many wrong and/or missing disks!

[Linkex]

Array has recently had new (re-used) disks installed, these had been built into the array and working for at least a couple of weeks. I do think this was the first reboot since then however.

 

In any case, rebooted the device and array won't start with error 'Too many wrong and/or missing disks!'. 2 disks are showing as wrong, however one of the 'new' discs replaced the parity device as it was larger, and that is showing green, so I don't THINK its just the new discs. showing as down

 

From what I've read, it looks like the best option is to create a new config, but I'm not clear on if that will impact existing data or not? Are there any other options I should be looking at?

unraid-diagnostics-20240104-2108.zip

[trurl]

Did you reassign the original parity disk as a data disk? Disk3 maybe?

 

That should have worked, but doesn't look like your changed assignments got saved to flash.

 

What do you get from command line with this?

ls boot

 

[Linkex]

I did reassign the old parity disk as a data disk, although I couldn't say for sure which slot it ended up in - I didn't realise it was something that should be tracked  Is there any way I can manually check which disk has parity data on it, as that would confirm if the disk currently in the parity slot is the actual parity disk?

 

'ls boot' gave an error, did you mean /boot?

 

root@Unraid:~# ls boot
/bin/ls: cannot access 'boot': No such file or directory
root@Unraid:~# ls /boot
EFI/                          bzimage           bzroot             changes.txt  license.txt          make_bootable_mac  syslinux.cfg-
System\ Volume\ Information/  bzimage.sha256    bzroot-gui         config/      logs/                memtest
bzfirmware                    bzmodules         bzroot-gui.sha256  ldlinux.c32  make_bootable.bat    previous/
bzfirmware.sha256             bzmodules.sha256  bzroot.sha256      ldlinux.sys  make_bootable_linux  syslinux/

 

[trurl]

yes, I was just checking to see if there was any FSCK files on flash that might have indicated corruption.

 

 

https://docs.unraid.net/unraid-os/manual/changing-the-flash-device/#what-to-do-if-you-have-no-backup-and-do-not-know-your-disk-assignments

 

[JorgeB]

11 hours ago, Linkex said:

From what I've read, it looks like the best option is to create a new config

A new config won't help in this case, this *should* work, type this for disks 3 and 4:

 

sgdisk -o -n 1:1M:0 /dev/sdX

 

Replace X with the wrong disks identifiers, disks 3 and 4, then reboot, disks will no longer show as wrong and the previous filesystem should mount.

 

 

[Linkex]

Thanks both for taking the time. I ran 'sgdisk -o -n 1:1M:0 /dev/sdX' on the disks, one is now showing green, but the other is still showing 'wrong'. It does now give me the option to start the array, but seems to suggest the disk will be overwritten if I do. Is there any way I can proceed without data loss?

unraid-diagnostics-20240105-1323.zip

[trurl]

You could try unassigning the disk to see if it is emulated by parity, but wait and see what @JorgeB says since he seems to have some idea of what is going on in this weird situation.

[Linkex]

I've just spotted, next to the start button it does say 'Start will start Parity-Sync and/or Data-Rebuild', so I'm quietly hoping that means it will rebuild disk 4 if I start the array. However, I will indeed wait for JorgeB to update before I touch anything, just in case there is a better/safer option 

[JorgeB]

1 hour ago, Linkex said:

but the other is still showing 'wrong'

Sorry, didn't notice it was a 2TB disk, that will be MBR, not GPT, unassign that disk and start the array, the emulated disk should mount and you can then rebuild, another way would be to get a 2TB disk MBR to replace that once, but since you don't have another 2TB disk a rebuild will probably be easier.

[Linkex]

Done, and it looks like my array is doing a data rebuild, so I do believe I'm now sorted. Thank you very much for that

 

Would I be right in thinking that this happened because I added disks with existing partition tables to the array? Is there something I could/should have done with the disks before adding them to the array that would prevent this?

[itimpi]

17 minutes ago, Linkex said:

array? Is there something I could/should have done with the disks before adding them to the array that would prevent this?

You could first delete any existing partitions.  This should be possible via Unassigned Devices.

 

alternatively running a Preclear on a drive will erase any existing partitions.

[JorgeB]

41 minutes ago, Linkex said:

Would I be right in thinking that this happened because I added disks with existing partition tables to the array?

Yes, but it's still a bug, it will be fixed for v6.12.7