smartctl shows Current_Pending_Sector 1, but smartctl -t log completes without error

[brettw.10]

Hi all,

My FreeNAS 11 RELEASE build is showing the following critical error:

Code:

Device: /dev/ada0, 1 Currently unreadable (pending) sectors

Following the recommendations from a number of threads, I ran a smartctl test:

Code:

smartctl -t long /dev/ada0

After the test was complete, I checked the status with smartctl:

Code:

root@thevault:~ # smartctl -a /dev/ada0
<snip>
=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED
<snip>
197 Current_Pending_Sector  0x0012   100   100   000	Old_age   Always	   -	   1
<snip>
SMART Self-test log structure revision number 1
Num  Test_Description	Status				  Remaining  LifeTime(hours)  LBA_of_first_error
# 1  Extended offline	Completed without error	   00%	 22268		 -

According to the other threads, there should be a value in LBA_of_first_error that you can then use to correct the issue. But, in my case, there is nothing, as the test passes without error. However, the Current_Pending_Sector never clears.

Anyone able to help me out here?

Rgds,
Brett.

 

[Johnnie Black]

Sometimes there are "false positives" on pending sectors, in my experience mostly but not limited to WD disks, likely a firmware issue, if the extended test passed disk is fine for now, but monitor for more issues, seeing the full SMART report might help to make a better assessment.

 

[Chris Moore]

You should schedule your SMART tests on all drives to run a short SMART daily and a long SMART test weekly. This is a preventive maintenance to allow you to detect a failing drive before it completely dies and leaves you in a bind.

Here is where it is discussed in the documentation:
http://doc.freenas.org/11/services.html?#s-m-a-r-t

Here is where they discuss configuring the scheduling:
http://doc.freenas.org/11/tasks.html?#s-m-a-r-t-tests

 

[brettw.10]

Running the risk of crossing streams (or threads) now. I was getting to the point where I wanted to start replacing the drives in this RAIDZ pool anyway. It currently consists of 4 x 1TB WD drives, but I am short on capacity. Is it possible to systematically replace each of the drives with a larger drive, allowing the data to be reconstructed on each replacement? Or is it best to back up all the data to an external drive, delete the pool, replace all the drives, and then bring the data back in?

 

[Ericloewe]

Is it possible to systematically replace each of the drives with a larger drive, allowing the data to be reconstructed on each replacement?

Yes, of course.

Or is it best to back up all the data to an external drive, delete the pool, replace all the drives, and then bring the data back in?

In your case, it's better because you could move to RAIDZ2.

 

[Chris Moore]

I was getting to the point where I wanted to start replacing the drives in this RAIDZ pool anyway.

RAIDz (RAIDz1) is not suggested for any drive larger than 1TB. The ideal situation would be to move the data, create a new RAIDz2 pool with 6 or 8 dries, then move the data back.