Hi guys,

 

my unraid box went down yesterday. Started looking into it this morning, and I'm a bit stumped. My feeling is it could either be an LSI controller or a cable gone bad (there's a bunch of drives missing from the LSi bios page), but the kernel panic message itself seems to point more towards a system thumbdrive issue, (it happens right after there's an attempted mount on it) even though it mounts and copies fine on an other machine.

 

Anybody got ideas ?

Shutdown your system and take out the USB stick. Let it do a file check and repair on a (Windows) machine.

 

1 hour ago, bonienl said:

Shutdown your system and take out the USB stick. Let it do a file check and repair on a (Windows) machine.

 

Well, wouldn't mount this time... but booted fine from a freshly flashed key. So I guess that's that... I'm a moron who didn't do a screenshot of my drive configuration, is that stored somewhere, or should I risk booting things up with the old super.dat ?

If you copy over the old super.dat file, it will preserve your current disk assignments, otherwise you will need to make new assignments

 

Also copy over your *.key file (found in the /config folder) to keep your existing license.

 

Just now, bonienl said:

If you copy over the old super.dat file, it will preserve your current disk assignments, otherwise you will need to make new assignments

 

Also copy over your *.key file (found in the /config folder) to keep your existing license.

 

 

Great, thanks a lot for the help !!!

Oookay... so, it looks like something happened, either from me hitting the power button, or from the usb drive going down.

 

Super.dat won't fly, and one of the drives shows as having no filesystem (I'd assume that was the parity drive, but I was running dual parity, so...). Is there a human-readable version of the previous drive assignments somewhere ?

2 minutes ago, mattbr said:

Is there a human-readable version of the previous drive assignments somewhere ?

Only if you made a screenshot previously.

 

You can do a "New Config" (see Tools) and go back to the Main page. Assign ALL your disks as DATA disks.

After starting the array, the disk(s) which show up as unmounted is (are) your parity disk(s).

Note these unmountable disk(s). Stop the array and now assign these disk(s) as parity (and parity 2).

Start again the array and a parity resync will commence... have patience to let it finish.

 

In case of doubt or you see something different/unexpected. Stop and ask before proceeding.

 

12 minutes ago, bonienl said:

Stop the array and now assign these disk(s) as parity (and parity 2).

Start again the array and a parity resync will commence...

 

There's a new config step before that, right ?

 

So new config - check which don't have a FS - new config - assign the drives without a fs to parity - start.

Yes, you are right. To assign parity disk(s) you will need to do again "New Config".

This time you can do "retain data disks" if you don't want to re-enter the data disk assignments.

Edited September 26, 20196 yr by bonienl

5 minutes ago, bonienl said:

Yes, you are right. To assign parity disk(s) you will need to do again "New Config".

This time you can do "retain data disks" if you don't want to re-enter the data disk assignments.

 

Figured - and I'd assume assigning the data drives randomly won't change a thing since parity is going to be rebuilt.

 

(in case someone else runs into something like this: remember to turn Docker off when you restart the array, in case one of the apps start hitting one of the drives. Looks like it happened here, so my (possible) pain is your gain)