I have a 3 TB WD Green disk that is a bit over two years old.  I happened to look at it in the unMenu myMain "Smart" view and it had a red highlighted value of

» current_pending_sector=133

Doing a bit of searching, I found this: http://lime-technology.com/wiki/index.php/Troubleshooting#Resolving_a_Pending_Sector

which says, in part,

The safest procedure is to replace the drive with a pre-cleared spare. The original drive can then be pre-cleared and the pending sector count should go to zero. The original drive can then be used as a spare. Multiple pre-clear cycles should not be required and the disk should be RMAed if 1 cycle doesn't work. If the drive cannot be returned then multiple cycles may restore the drive to a usable state

Now this disk had a previous write error and red-balled.  I replaced it with a new disk and then ran preclear on it for 3 cycles.

Here is the smart data after 3 cycles:

 

** Changed attributes in files: /tmp/smart_start_sdl  /tmp/smart_finish_sdl

                ATTRIBUTE  NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS      RAW_VALUE

          Seek_Error_Rate =  100    199            0        ok          0

      Temperature_Celsius =  119    122            0        ok          31

  Current_Pending_Sector =  200    198            0        ok          0

No SMART attributes are FAILING_NOW

 

1221 sectors were pending re-allocation before the start of the preclear.

0 sectors were pending re-allocation after zero of disk in cycle 1 of 3.

0 sectors were pending re-allocation after post-read in cycle 1 of 3.

0 sectors were pending re-allocation after zero of disk in cycle 2 of 3.

0 sectors were pending re-allocation after post-read in cycle 2 of 3.

0 sectors were pending re-allocation after zero of disk in cycle 3 of 3.

0 sectors are pending re-allocation at the end of the preclear,

    a change of -1221 in the number of sectors pending re-allocation.

7 sectors had been re-allocated before the start of the preclear.

7 sectors are re-allocated at the end of the preclear,

    the number of sectors re-allocated did not change.

 

Now, about 2 months after I put it back into service, the Current_Pending_Sector count is 133.

Question is: should I bother to try to clear this by writing or should I attempt to RMA it?

It is in warranty but I don't know if WD considers this type of "error" to be reason for RMA.  They only ask if there are "bad sectors".

 

Anyone who has had this situation in the past care to comment?

Thanks.

 

 

RMA.

Pending reallocated sectors can completely corrupt and invalidate your parity and data.

RMA.

Pending reallocated sectors can completely corrupt and invalidate your parity and data.

 

Thanks.  I had planned to use this technique : http://lime-technology.com/forum/index.php?topic=37431.msg346187#msg346187

to write zeroes to the disk and then remove it from the array.  This is to prevent having to rebuild the parity and be exposed during that time.  However, I don't know if this will work on a drive that has pending sectors.  If you believe this:

The clearing step forces the SMART firmware to resolve all marginal sectors, either recover them or remap them, and should result in clearing the Current_Pending_Sector count to zero.

from this link: http://lime-technology.com/wiki/index.php/UnRAID_Add_Ons#Preclear_Disk

then the dd write of zeroes will cause the pending sectors to be reallocated.

 

Any thoughts on this?

 

Your pending sector count seems to be continually increasing. This behavior indicates a failing drive that should be replaced. If the drive is under warranty then replace it ASAP. Getting the pending sector count back to zero doesn't accomplish anything if the pending sectors just keep coming back.

The disk has had this issue twice and should be RMAed or discarded.

RMA.

Pending reallocated sectors can completely corrupt and invalidate your parity and data.

 

That's not likely -- but what they CAN do is make it impossible to successfully rebuild a different disk.

 

A pending sector is a sector that the disk's SMART system knows is bad; but can't be reallocated because it hasn't been able to successfully read the data to move to a reallocated sector.    If you access that sector and get a good read, it will be reallocated;  or if you rewrite the sector, it will be reallocated since SMART then knows the old data doesn't matter.

 

What's more likely in typical operation is that when you try to read a pending sector, the disk tries a few times, then returns an error -- and UnRAID simply emulates the data by reading all of the other disks and the parity drive.  I believe that in this case UnRAID also rewrites the sector with the good data ... which will force a reallocation of that specific sector.  If you had a utility to read every sector and rewrite it on a disk with pending sectors, the end result would be no more pending sectors, as they would all be reallocated.    This is, of course, what happens if you simply write zeroes to the entire disk (i.e. run a pre-clear) ... but it's quite possible to do this without any data loss by simply rewriting every sector with its current data (emulated data for those that return errors).

 

But if a DIFFERENT disk was to fail, then when you attempted to rebuild it, the data from the pending sectors would still not be readable, so the attempt to rebuild the other disk would likely fail.

 

A simple (although time-consuming) way to clear pending sectors is to force a rebuild of a disk onto itself.    Stop the array; Unassign the disk; then Start the array so the disk is shown as "missing";  then Stop the array again; assign the disk back to where it had been; and then Start the array and it will do a rebuild.    This clears all pending sectors (since they'll be reallocated).

 

Yeah, if you cannot rebuild another disk then parity data is corrupted / invalidated.