Cannot boot from usb after reboot (did stress test before)

[Aldamir]

Hi,

i can't boot from my unraid USB anymore after i started a prime95 and fios stress test (

 

There were some weird signs when i was still in the system (docker service couldn't be found according to UI and then the array was stopping). After some time the UI wasn't reachable also nothing via SSH. I'm not sure if this issue is related to the tests thought, because i changed some settings in the Tipps and Tweaks plugin afterwards and i couldn't see any issue in using the interface at first glance.

 

Then I tried to restart, but it won't boot to the USB anymore; also tried changing the USB port, toggling the fast boot and secure boot settings.. nothing.

 

In the connect-tab by unraid, I can't see any flash backups so i'm pretty desperate and would really appreciate any help!

Edited September 15, 2024 by Aldamir

[itimpi]

Have you tried plugging your flash drive into a PC/Msc and seeing if you can read its contents?   The contents of the ‘config’ folder is all that is needed as that is where everything specific to your server is stored.

 

There are ways of recovering if you cannot read the current flash drive contents but they are less convenient.

[Aldamir]

Yes i tried and then executed the make bootable script as suggested by someone on another forum. After that, the flash won't mount under mac or windows. On linux, i can see the content, but there are some errors when copying plugins.. 138 of 149 mb are copied. Is there any way to restore the basic settings at least? I mean i have a backup or rather a snapshot of the appdata on my array..
Thanks a lot!

[JorgeB]

Do you have a recent flash drive backup?

[Aldamir]

I thought, that i would have one, but my connect section in myunraid seems to be empty...

But i think that most of the config folder should be fine, the errors where only on the plugins folder. Can i create a clean flash drive and just put everything on there except the plugin files?
The most important would be the docker configuration, and i did a appdata backup with the Appdata Backup Plugin.

Thanks a lot!

[itimpi]

1 minute ago, Aldamir said:

The most important would be the docker configuration, and i did a appdata backup with the Appdata Backup Plugin

Unfortunately the docker configuration files are stored under the plugin folder.   Can you get at any of the ‘plugins’ folder contents?   The docker config files are stored in the plugins folder under dockerman/templates-user.

[Aldamir]

Omg, yes - those seem to be intact! 😍 We are talking about those XML-files, right?

[itimpi]

Just now, Aldamir said:

Omg, yes - those seem to be intact! 😍 We are talking about those XML-files, right?

Yes

[Aldamir]

Very nice, this kind of worked, i copied the docker folder to the new config file, replace the existing GUID with the new for my key. Docker are up and running.

 

But one of the drives is reported as unmountable, is there maybe some mismatch in the slots of the array? Can i somehow get the previous array slots?
Thanks a lot!!

[JorgeB]

12 minutes ago, Aldamir said:

But one of the drives is reported as unmountable

Please post the diagnostics.

[Aldamir]

Logs are attached, please don't mind the key error - i had to reassign the key to the new flash drive - thanks for checking and helping!

thonas-diagnostics-20240916-1530.zip

[JorgeB]

Check filesystem on disks 1 and 2, run it without -n, disk1 is mounting but it's also showing issues.

[Aldamir]

Checked the filesystem and it seemed to do the trick - the Disk 1 (the one that actually mounted) reported a fatal error during the process with the comment to rerun the check, the second run had the attached output. I think some files were put in a lost+found share.

 

I'm not entirely sure but i think the connect plattform showed me 75% usage of the array and now only 70% - is it possible that some data was completely lost apart from the lost & found?

 

Should i run a parity check now? Thanks for the quick help!

 

Disk2_sdc_output.rtf Disk1_sdd_output.rtf thonas-diagnostics-20240916-1530.zip

Edited September 16, 2024 by Aldamir

[JorgeB]

24 minutes ago, Aldamir said:

I'm not entirely sure but i think the connect plattform showed me 75% usage of the array and now only 70%

Was that when the disk was unmountable? It can show more used space since there will also be less capacity, but it's possible that the repair loses some data, it will depend on how bad the corruption was.

 

 

 

 

[Aldamir]

This was when the disk was unmountable, yes! But since i only have three disks, one of which parity, it should show a bigger difference, i guess?

But i guess, i would have to live with the loss of data anyway

 

Since the disk is mountable again, should i start the parity check now? Thanks for all the support!

[JorgeB]

56 minutes ago, Aldamir said:

But since i only have three disks, one of which parity, it should show a bigger difference, i guess?

The fewer the disks, the more difference it can show.

 

Say you have:

disk1 10tb - 10TB used

disk2 10tb - 5TB used

disk3 10TB - 5TB used

 

if disk3 is unmountable, you will have 75% total space used (15TB of 20TB), if disk3 is mountable, it will show 66% total used (20TB of 30TB)

[Aldamir]

Ah yeah, makes sense! I only have three drives (2 data drives, 1 parity) and i double checked now. The mounted disk is used with 75%, which should explain why it showed it this way. Thanks for explainig

 

So can i start the parity check now, right?

[JorgeB]

You, but parity should be in sync.

[Aldamir]

Ok thanks a lot for all your help @JorgeB @itimpi - would have struggled a lot longer or not be able to rescue my stuff if you didn't help!!