The situation is that a drive showed up as failed, and was being emulated.

 

Shutting down the array and running an extended SMART test on the drive it actually seems fine, however my partiy drive that unraid says is fine just isn't.

 

When restarting the array for a rebuild the parity drives is all read errors after a few minutes, and won't even finish a short SMART test, never mind an extended one.

 

6 drives ( 5 + 1 )

 

My understanding is that I'm better off just starting a new array without the parity drive, the original failed drive will still have it's data unless formatted, and then install a new parity.

 

Before I do anything, I wanted to check if that is correct?

 

Thanks

Edited May 30, 20233 yr by dsmith44

Please post the diagnostics.

unraid-diagnostics-20230531-1022.zip

SMART for disk2 is also showing some issues, and there's a recent failed SMART test, though the last one passed, so with some luck doing a new config a rebuilding parity should work.

Thanks

 

I am slightly lost on best way forward here.

 

The array thinks the parity is fine, but we know it isn't, it is effectively dead, meaning the current array is toast given that it thinks the other drive isn't ok.

 

I think what I need to do is start again, but absolutely do not want to loose my mirrored cache pool.

 

To do that I plan to do the following

New Config

Preserve all slots

Remove current failed parity from new config.

Config will be 1+1 SSD cache pool, 4 array drives without a parity (disk0/sdj) , disk2/sde questionable.

Start array.

Move data off potentially failed disk2 to disk1/3/4.

Shutdown array.

 

Add 2x6tb new drives, probably after a zero to make sure they are ok.

Add one as new parity.

Add 2nd as new drive.

Remove disk2.

 

Start array and wait for parity to build.

 

Is that my best way forward?

 

Thanks

You could also just do a new config and add parity now and see if it builds without errors, if it does replace the disk after.