Hello,

I have a question about how Unraid handles uncorrectable read errors on a data disk.

I recently saw the following SMART error on one of my HDDs:

• SMART Error: UNC (Uncorrectable Error)

• Error occurred during a normal read (READ FPDMA QUEUED)

• Specific LBA reported as unreadable

My question is about Unraid’s behavior in this situation:

1. When Unraid encounters an uncorrectable read error on a data disk:

   • Does Unraid use parity to reconstruct the missing data in memory?

   • Is the reconstructed data then written back to the same disk (potentially forcing sector reallocation)?

   • Or is the disk only marked as having a read error without attempting a rewrite?

2. Under what conditions will Unraid:

   • Automatically reconstruct and rewrite the bad sector?

   • Disable the disk?

   • Leave the sector pending until a manual operation is done?

regards.

It uses parity to reconstruct the data and writes that block back to the disk, if it succeeds, nothing else is done; if the write fails, the disk is disabled.

thanks !