I've searched but cannot find any clear concise instruction to remove a failed data disk from the array and allow remaining good data disks to rewrite the data from the failed disk.

 

I received alerts that disk6 was overheating (63°C) two days ago but still passed a smart test.  I stopped the array and rebooted unRAID, array is still stopped.  Disk6 is disabled.

 

Do I physically remove disk6, start the array and run a parity check to rebuild disk6 data to the remaining data disks?

You must rebuild onto another disk. There is no way to do what you are asking. It might be possible to copy the data from the failed disk, or from the simulated disk, but you would have to do it manually and you would be completely unprotected until you rebuilt parity without the disk.

 

Best approach would be to preclear a new disk and rebuild onto it. Preclear is not strictly necessary for a rebuild, but it would be best to test the new disk before trusting it in your array.

And you can only rebuild to a disk as large as the original disk, but no larger than parity (unless using the parity swap procedure).

 

Since you are unclear how parity works, you should proceed carefully and ask for additional advice. If you make a mistake you will lose data.

Thanks trurl.  I'll leave the server off and purchase a new disk equivalent to the current parity disk (4TB) and rebuild with the new disk.

 

Update:  I installed a new 4TB disk, rebuild missing disk then performed parity check, zero errors/data loss.