Hey guys,

 

I'm new to unRAID and so far, I've managed to get my system setup pretty easily. But one thing that occurred recently is that unRAID threw the following error one day:

 

Event: unRAID Disk 1 SMART health [197]

Subject: Warning [uNRAID] - current pending sector is 8

Description: WDC_WD80EFZX-68UW8N0_VK0R3V0Y (sdi)

Importance: warning

 

I did a extended self-test, but there was nothing wrong:

Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error

# 1  Extended offline    Completed without error      00%      314        -

 

I then pulled the drive and plugged it into another PC and ran an extended test with the WD Lifeguard Diagnostic...and there were no errors either.

 

However, the disk attributes still show an error:

 

Apart from my OCD not being able to stand that one orange line, is it something I should be worried about? Given that it passed the WD Diagnostic with no errors, I'm not even sure I will be able to RMA it either.

 

Hope you guys can advise! Thanks!

WD disks are prone to these false positives, disk is fine for now, although previous pending sectors are never a good sign, keep an eye on it.

The pending sectors indicate that the SMART system has identified a sector as marginal during a read (however it makes that determination is manufacturer or even model specific).

 

When a pending sector is WRITTEN, the SMART system will do another determination on that sector, and decide whether to remap it or UNmark it as pending.

 

So, I believe that diagnostics that don't write to the disk will not clear this pending sectors. And each of the reads to these sectors may have been shaky, but if they were successful, it might very well report all is fine.

 

I would want to preclear the disk. Preclear in particular will read the disk (think of it as cocking triggers on sectors that might be marginal), write the disk (gives disks chance chance to pull the trigger to remap bad/marginal sectors), and then read again (cocking additional triggers). Generally if you get through all of that with no sectors getting marked, all is good. If the initial read marks something (or something was already marked pending), that is not a death knoll, but is certainly indicative of a drive to keep your eye on. But if you come out of the post read with more sectors marked, that is bad. We just read and wrote and read, it should be solid. Truth is when disks start to go bad, we start to see changes in reallocated and pending sectors on every pass through the disk. This points towards a failing drive.

 

But I will say that all software has bugs. And SMART systems in drives (which few people even know exist much less focus on) are likely a prime candidate. I have seen pending sectors that never get remapped and never cause a problem. They just stay put. So if I had a drive go through a preclear and set of diagnostic tests and the pending and reallocated sectors hold steady, I'd tend to trust that disk until I start to see changes. It is really the changes in these values that matter more than the absolute number.

In my experience pending sectors can't be read, if the disk passes an extended SMART test there are no pending sectors, this happens a lot with WD disks where I believe the sectors are reallocated fut the SMART is not correctly updated, you can run 10 preclears and SMART will still say that there are pending sectors.