SMART error showing on the parity disk

[Loosetooth]

Hello.

First off I should apologise.  I am a bit of a noob and I have tried searching the forums for an answer to this... so, I know this is basic.  But I am not finding an easy solution on my own.

A couple of years ago a friend encouraged me to build a basic server into a HP Microserver and, at the same time, I built one for my father.  Well, I never finished the project before the pandemic hit and I have lost contact with that friend.  So, these servers have lain idle for a couple of years.

I used WD 4tb Purple drives and they were bought new.

Now, I am upgrading my fathers PC and he wants to get his server up an running and so I am trying to get a basic understanding of Unraid and get these servers going.

Now... On my fathers server the parity drive is displaying an SMART error.  I have run self-tests... I have tried reseating it and cleaning the contacts.  It looks and behaves like a new drive, I have tried reading the attribute page, etc.  I dunno whether I should try to reformat it (although I cannot find how to do that... in the forums the displayed screenshots seem to have different buttons to me)... And I dunno whether to replace the drive and not use it (it being my fathers server I do not want to make a mistake and ingrain a drive failure).

If the drive is faulty, I will swap it out... but it should not be (in an ideal world) and whatever ails it... it does not seem to be a big thing, so if I can fix the error it would provide some peace of mind.

Again, I know that this is a little basic and so I thank anyone/everyone for their time and replies.

[JorgeB]

Please post the diagnostics.

[Loosetooth]

Hiya, I have read the link and assume you want the zip file?

Many thanks

eaufarm-diagnostics-20221102-1137.zip

[JorgeB]

The drive failed the extended SMART test, I would let it sync parity, pending sectors should go to zero when done, then run an extended SMART test, or a non correcting parity check.

[Loosetooth]

Hiya,

Well... I ran the extended SMART test last night in response to it showing an error.

 

When you say "sync parity", how would I do that?  By starting the array?  (I did not start it cos I had a disk showing an error) or is there a link to a webpage that will give me instructions on how to proceed.  I really want to master small servers and Unraid, but right now I feel a bit all at sea.

Thanks for your advice.

[JorgeB]

6 minutes ago, Loosetooth said:

By starting the array?

Yep, that will start the parity sync.

[Loosetooth]

Lovely, thank you

[Loosetooth]

22 hours ago, JorgeB said:

Yep, that will start the parity sync.

Hiya,

So... I started the Parity Sync, which was all successful.  Then I formatted the remaining drives and ran another SMART extended self-test.

Still came back with one error on the "Attributes" tab and shows it is a drive with errors on the dashboard.

Should I replace the drive to be safe?
 

[JorgeB]

Up to you, should be OK for now, I would give a 2nd chance but if any more errors appear then replace it.

[Loosetooth]

3 minutes ago, JorgeB said:

Up to you, should be OK for now, I would give a 2nd chance but if any more errors appear then replace it.

 

I think the problem is that this is the server for my father.  He wants something to back his photography up to.  The disk displaying a SMART error is a visually discomforting thing.  Obviously because these are "new" disks (2 years old, but not been used for anything) it should not have any errors.  So, how important are SMART errors?  Like, is this server actually fine and dandy, but just has a single bad sector and should be good to go?  Or is the error a signal that there is something wrong on the drive that could, potentially, lead to my father losing data later?

 

[JorgeB]

It it still showing pending sectors? It should have gone to zero after the sync.

[Loosetooth]

So, the pending sector errors are gone, as you described.  The one that has remained is the #187 reported incorrect error.  This has a RAW value of 1.

 

https://www.backblaze.com/blog/what-smart-stats-indicate-hard-drive-failures/

 

I am reading up trying to understand what it all means... and the guy above talks about it.  There is one side of my mind which says... it appears to be a singular error and that in and of itself it is probably one of those things and it is fine.  The other side of my mind says... well, this is the parity drive (is it more serious if the parity drive fails?), it is not for me but a close relative, they are storing data which is highly precious to them, it could be an indicator of future trouble, ideally you would want to start of with no errors and it might be easier to nip it in the bud now.

But it is hard for me to make a call, because the word "error" is, due to my ignorance of servers and Unraid and SMART, highly emotive.  This might be nothing to worry about.

[JorgeB]

That would be OK for now, as long as an extended SMART test or parity check completes without any errors.

[Loosetooth]

The last extended test completed with read failures, but the parity completed and is valid. 

[Loosetooth]

I will start a parity check but it will take 6 hours

[JorgeB]

11 minutes ago, Loosetooth said:

The last extended test completed with read failures

This was before the parity sync correct?

[Loosetooth]

1 minute ago, JorgeB said:

This was before the parity sync correct?

Nope, after.

So, on the back of our discussion I started the array, formated the drives and then did an extended test.  It returned with read failure.

[Loosetooth]

[JorgeB]

45 minutes ago, Loosetooth said:

Nope, after.

In that case the disk should be replaced.

[Loosetooth]

21 minutes ago, JorgeB said:

In that case the disk should be replaced.

Cool,

Thanks Jorge.

[Loosetooth]

Crikey... I am still having problems.

I swapped out the parity drive with SMART error showing... and replaced it with a drive of exactly the same type and age. 

I did not know if there was a certain protocol or way that I should do that... so I took the array offline, turned the server off.  Replaced drive and then turned everything back on.  On the "MAIN" page of Unraid I then assigned the new drive, in the drop down menu, as the new parity drive.  I then hit "START" so that it would do a parity check and rebuild the array.

 

This it did overnight.

 

I now have green everywhere on the dashboard.... but...

When I run a SMART self test it comes back as "READ FAILURE".  But it does not turn amber.  Extended tests complete almost instantly... and short tests run for a while, but then return the same result.  But it all comes back green, ie it does not turn amber.

Is this another bad drive, or something else at play here?

 

 

 

[JorgeB]

Post new diags.

[Loosetooth]

11 minutes ago, JorgeB said:

Post new diags.

 

I actually dragged that drive out and replaced it again (I have two microservers and 8 drives, now 6).  I SMART tested the third drive and it came back sound and am rebuilding/doing parity sync with the third drive.  Will the diags still be relevant or have I effectively wiped them?

eaufarm-diagnostics-20221104-1100.zip

[JorgeB]

If it's a different drive there won't be a SMART report, but looks like the other one was also failing.

[Loosetooth]

15 minutes ago, JorgeB said:

If it's a different drive there won't be a SMART report, but looks like the other one was also failing.

🙄😒

 

My luck continues.  It is a pity, these were good drives and bought new... but they have been lying dormant for a while.  I will try and have a run at them with some software and see what is what.

Ta, again, Jorge.