Inconsistent SMART results?

[WiseHacker]

Like many others, I'm getting SMART errors. Before I return this drive though, hopefully I can get some more insight from here as my SMART results seem to contradict themselves.

In short, I'm in the process of building a new server (even have a TrueNAS system on the way). I just got some new hard drives and the first thing I did was run an extended test (literally what I do for any drive for a NAS).

It's bad enough the extended test fails with a read error near the end but it only gets confusing (for me) from there.

Here is what SMART is reporting for measures 5, 197, and 198.

Code:

  5 Reallocated_Sector_Ct   0x0033   100   100   010    Pre-fail  Always       -       0
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       8
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       8

If I understand this right, the 8 in 197 and 198 effectively mean the drive is bad and I should return it.

But it's measure 5 that confuses me. I have 8 sectors pending but no reallocation events?

The final bit of frustration is the log summary below. I ran a short test for laughs, but what is important is the result of the long tests.

Code:

SMART Self-test log structure revision number 1
Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
# 1  Extended offline    Completed: read failure       20%        19         -
# 2  Short offline       Completed without error       00%        11         -
# 3  Extended offline    Completed: read failure       20%         8         -

Seriously, where is the LBA of the first error? If I have bad sectors, I should at least get the LBA of the first error, right?

Is this drive honestly bad? Or am I about to prematurely return a workable drive?

Any help understanding this would be great as I prefer my SMART report to literally say "all good" or "He's flatlined, Jim". All the other threads on the matter show an LBA yet I am not.

My thanks in advance.

 

[joeschmuck]

But it's measure 5 that confuses me. I have 8 sectors pending but no reallocation events?

Pending means there was an error reading but it eventually read the data, so it's not been actually reallocated yet. ID 5 is when it fails enough times.

Seriously, where is the LBA of the first error? If I have bad sectors, I should at least get the LBA of the first error, right?

You are correct, the report you listed does not indicate the failed LBA, so use the "-x" switch vice "-a" switch with smartctl to get the full list from the drive. Hopefully it has the data.

Any help understanding this would be great as I prefer my SMART report to literally say "all good" or "He's flatlined, Jim".

The SMART Status is rarely correct when you have surface errors like it appears you do. Your drive is bad and you should back up any data you desire to keep and replace the drive.

In short, I'm in the process of building a new server (even have a TrueNAS system on the way). I just got some new hard drives

Are they new to you? White Label?, and also you should post the make/model, and you should look up the serial number for the manufacturer to see if they are still under warranty.

Good Luck!

 

[WiseHacker]

Thank you, joeschmuck.

When I say new, I mean new-new not new-to-me, sorry.

It's so new, it has not even been formatted for use with ZFS so it literally has zero data for me to back up on it.

I also apologise, I did run smartctl with -x, and completely missed the list in it.

For reference here it is below:

Code:

Pending Defects log (GP Log 0x0c)
Index                LBA    Hours
    0        12776698368        8
    1        12776698369        8
    2        12776698370        8
    3        12776698371        8
    4        12776698372        8
    5        12776698373        8
    6        12776698374        8
    7        12776698375        8

I was already planning to replace, but some searches online hinted that the drive could still be put to use, if I could force the thing to reallocate those sectors first.

Given I still have a pending list, is it safe to say the drive has no reserve sectors to replace/re-map the bad sectors?

I know all drives these days have a reserve as sectors go bad, but when they show in SMART (if I understand right) it's supposed to mean the reserve have been depleted.

 

[joeschmuck]

but some searches online hinted that the drive could still be put to use, if I could force the thing to reallocate those sectors first.

While you could force the bad sectors to be marked as bad, normally if the plater media is flaking off (the typical reason for this kind of error) then it would likely continue for a while. It is much smarter to replace the drives, especially since they are new.

When I say new, I mean new-new not new-to-me, sorry.

Do not apologize, I asked that because often someone will buy drives on EBay as slightly used or "new pull", or even White Label which often means no or very limited warranty.

It happens that a bad drive finds it's way out of the factory or is damaged during shipping.

You should be happy that you were running an extended test, a lot of people do not and the problem will show up months later. Also I would return it to the place you purchased it, do not use the warranty service unless you have to.

Take care.

 

[WiseHacker]

You should be happy that you were running an extended test, a lot of people do not and the problem will show up months later.

Let's just say I had an incident with WD Red back in 2019. This was before the SMR uproar.

I bought five 8 TB drives (one was to keep as a spare).

One of them was bad so I had it replaced. And the replacement was even worse so I simply got a refund.

Since then, I do an extended test first on any drive I put in a NAS and even before formatting it.

Also I would return it to the place you purchased it, do not use the warranty service unless you have to.

It's OK. I'm in Australia so the consumer does have more options when it comes to warranty.

Just filled in the forms for an RMA now so come Monday hopefully I will get the needed authorisation.

Fingers crossed the replacement is OK. I'm only going with Ironwolves now because WD Reds are just a royal mess now.

 

[NugentS]

just to agree with other posters. That drive is unacceptable as a new drive. Return it and buy another.

As an aside I run a destructive badblocks test on any new HDD before I use it. This test can take a week to complete - but its a good stress test of a new or "new" drive before using it