I recently noticed that one of my drives had a current pending sector count of 1.  I ran an extended SMART test, nothing changed.  I ran a parity sync, had no errors.  I had enough free space to move all of the data from that drive to others and remove the drive from the array entirely.  I then ran preclear on the drive, thinking that it would force the pending sector to be remapped, however it remained at 1 following the preclear (see the attached SMART report).  No other indications of problems with the drive.

 

Is it safe to put the drive back in the array?  It is still under warranty, so I can also RMA the drive and get a replacement. 

 

preclear_start__WD-WCAVY2438779_2013-02-23.txt

preclear_finish__WD-WCAVY2438779_2013-02-23.txt

You included the start and finish SMART reports, but you did not attach the actual preclear report.

 

It would show when the sector was marked as un-readable.    Typically, we see it re-mapped after zeroing the drive, and detected AGAIN as un-readable in the post-read phase.

 

Joe L.

Sorry about that, the preclear report is attached.  The sequence of events that got me here was:

 

1) Power failure (wife didn't think a UPS was important...changed her mind now).

2) Errors in resulting parity sync caused me to run SMART tests on all drives.

3) Identified pending sector on 1 drive, ran an extended test, no change in the report.

4) Ran another parity sync, all was fine.

5) Ran more smart tests on all drives, no changes.

6) Copied all data from the bad drive to others.

7) Another parity sync, all was fine.

SMART tests on all drives, no changes, the one drive still had 1 pending sector.

9) Removed the drive from the array, ran preclear.

 

Note that I ran preclear with the -W option because I had run a parity sync with the drive installed in the array immediately before removing the drive from the array and preclearing it.

 

preclear_rpt__WD-WCAVY2438779_2013-02-23.txt

I have no idea.  It sure is not giving any clues where that one sector might be.

 

You might try a "long" test.  It might give you the address of the pending sector.

Disable any spin-down timer on the disk, then type:

smartctl -t long /dev/sdX

 

Then wait 4 or 5 hours and get another SMART report and see if the long test provided any detail.

Long SMART test finally finished after about 7 hours, unfortunately it gave no further information (report is attached).

 

sdc_smart.txt

Attach the drive to a Windows PC and run the Western digital diagnostic tools on it, this has sometimes found different drive problems for me.

 

Regards

Stephen

Long SMART test finally finished after about 7 hours, unfortunately it gave no further information (report is attached).

That is probably good news...  (the counter might be wrong, and the disk has all readable sectors, as the long test proved)