Erros on disk 8

[neuk34]

Hello,

 

I went back from vacation, I switch my server on and disk 4 went off.

A parity check has been automatically launched and discover 256 errors on disk 8 (xfs).

I've ordered a new drive but I want to be sure that nothing will go wrong.

Could you please help me to avoid any data lost?

 

Thanks

 

nas-diagnostics-20200828-0950.zip

[JorgeB]

Disk8 appears to be failing, and since there's already a disabled disk and you only have single parity you can't rebuilt it, you can try copying everything you can from that disk to others or using ddrescue.

 

Disk4 is not showing SMART, you'll need to check connections.

[JorgeB]

Forgot to mention, CPU is overheating, check/clean cooler.

[neuk34]

1 hour ago, johnnie.black said:

Disk8 appears to be failing, and since there's already a disabled disk and you only have single parity you can't rebuilt it, you can try copying everything you can from that disk to others or using ddrescue.

 

Disk4 is not showing SMART, you'll need to check connections.

you'll find smart report here.

thank you

nas-smart-20200828-1306.zip

[JorgeB]

SMART isn't working.

[neuk34]

In this situation, does it mean that I won't be able to recover disk 4 datas ?

[JorgeB]

Possibly if disk4 is really failing, try power cycling the server and then manually get a SMART report, post the output of:

 

smartctl -x /dev/sdX

 

Replace X with correct letter, it might change after rebooting.

[neuk34]

@johnnie.black, I decided to clean my server when you mentionned that CPU was overheating. A lot of dust has been removed.

 

Disk4 is clearly dead, but content is still emulated thanks to parity drive.

I've plugged a fresh new 4 To drive that I added to the array.

This new drive (disk12) has been automatically cleaned up and formatted into Reiserfs.

Parity check has not detected any error, even on disk8 (that previously shown 256 errors - maybe overheating with dust?)

 

Then, I started to copy the emulated content of the failed disk4 directly into this new disk via terminal (mc with RenMov).

After 180 Go transferred, something went wrong and array freezed.

I had to make an uncleaned shutdown to restart server.

Now, my new drive is stuck at "mounting". New uncleaned shutdown to restart.

 

I started array into safe mode, but it still stuck at mounting status.

I also tried to reiserfsck  disk12 without success.

 

What should I do ?

 

thanks

nas-diagnostics-20200901-1523.zip

Edited September 1, 2020 by neuk34

[JorgeB]

Reiserfs can take a long time (hours) to mount large disks after an unclean shutdown, just wait, if there's a HDD led activity it should be on.

[neuk34]

Thanks @johnnie.black After few hours, my array is now alive as you predicted.

I launched a new parity check, and my disk 8 is unfortunately still showing 128 errors, with disk4 disabled.

Do you think that I should be able to rebuild disk4 if I plug a new fresh disk replacement, or it won't work?

 

Should I :

- transfer emulated content through terminal from parity -disk4 to disk12

- buy a new fresh disk to recover disk8 datas thanks to ddrescue

 

what do you recommend?

[JorgeB]

You have two failing disks with single parity, copy everything you can from emulated disk4 and disk8 to other disks in the array, then do a new config with the remaining good disk and re-sync parity, alternatively you can use ddrescue on disk8 and actual disk4 then do a new config with the clones.

[neuk34]

Everything went fine ! I recovered all my datas so thank you.

Last point : do you know how to format an unassigned device ?

I follow Settings > Unassigned Devices > Destructive Mode and set it to Enabled but it doesn't work.

 

Thanks

[JorgeB]

See the first couple of posts in the UD plugin support thread:

 

[neuk34]

Works great ! thanks