Jump to content

Preclear result interpretation


boosted

Recommended Posts

** Changed attributes in files: /tmp/smart_start_sdh  /tmp/smart_finish_sdh

                ATTRIBUTE   NEW_VAL OLD_VAL FAILURE_THRESHOLD STATUS      RAW_VALUE

   Throughput_Performance =   128     100           54        ok          116

      Temperature_Celsius =   216     232            0        ok          30

 No SMART attributes are FAILING_NOW

 

 16 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 1.

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

    a change of -16 in the number of sectors pending re-allocation.

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

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

    a change of 2 in the number of sectors re-allocated. 

 

So I did a preclear on a new 8TB WD Red.  This is what I see after the pre-clear.  I understand re-allocations are fine. But the 16 pending that went away seems weird.  Why would it go away and not be re-allocated?  Since the drives are still new, would it be best to get it exchanged before I have to get it warrantied? (warranty = refurb drive, exchanged = new drive)

Link to comment
4 hours ago, boosted said:

So I did a preclear on a new 8TB WD Red.  This is what I see after the pre-clear.  I understand re-allocations are fine. But the 16 pending that went away seems weird.  Why would it go away and not be re-allocated?

 

Because there need not bee anything wrong with the physical sectors. It might just be that the old content couldn't be read correctly - but after having been rewritten the new content can be read back correctly.

 

If someone smudges the text on a white board - is it the board that needs to be replaced or is it enough to just rewrite the text?

 

Next thing is that there are multiple reasons why the data couldn't be read in the first place. It just could be that the sectors are slightly weak but it's quite common that it could be because there was some interference when the data was written - possibly vibrations. The disk itself doesn't know if a write goes well until it tries to read - the writes are performed "blind". When the disk wants to write a sector, it first uses a read head to seek to the correct track. Then it uses the read head to follow the track while waiting until it gets near the intended sectors. Then the drive switches from the read head to the write head - and since the read and write heads are at different locations on the swinging arm, the drive has to make a small realign of the heads when switching between read and write mode. It does this without any feedback. Then the drive counts time until the intended sector rotates in under the write head before the drive starts to write.


If you have vibrations, then it's possible that the write head ends up being slightly unaligned - the drive will not know that while performing the write, unless the vibrations were strong enough and it's a drive that happens to have a vibration sensor (something enterprise disks and most NAS disks have) so it can abort and then retry the write. After the write, the drive can make a quick check that it's still aligned on the correct track - but it normally can't take the time to wait a full revolution to go back and try to reread the just written data. So vibrations can then result in a tiny fraction of written sectors to be hard to read - and sometimes impossible to read.

 

So vibrations are bad. And drives that doesn't have vibration sensors shouldn't be used in multi-disk installations.

 

And in summary - offline uncorrectable doesn't mean the physical drive is broken and that the sectors needs to be replaced. Just that the drive can't read and correct the original data and rewrite it. So it needs to wait for the computer to try to write completely new data before it can attempt to write to the problematic sectors.

Link to comment

Thank you for the detailed explanation.  Since this is the first disk to even have any reallocated sectors, I'll run another pre-clear to see what might show up.  If not, I will put it into service.  Unfortunately for consumer stuff, we can't have vibration sensors and have it cheap.  

Link to comment

maybe there is a problem.  After the initial preclear, I ran 2 more without a pre-read, then ran a 3rd with a pre-read.  The 2 without pre-read came back clean, but the 3rd with a pre-read now reads:

 

No SMART attributes are FAILING_NOW

 

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

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

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

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

    a change of 8 in the number of sectors pending re-allocation.

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

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

    the number of sectors re-allocated did not change.

 

more sectors in pending.  

Link to comment

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...