Hopefully this isn't too long, but in my last parity check (last week), I received 493 read errors on drive5 (sdl) in my array.  Using the unbalance plugin, I moved all of the data to drive9 (sde) in the array.  The read errors increased to nearly 1000 (perhaps doubled??), but I can't recall the exact number.  Anyway, I just figured it ran across the same read errors when unbalance was moving the data to drive9.  I didn't see any errors in unbalance indicating it couldn't move the files.

 

I tried to run Fix Common Problems, but it hanged and wouldn't start.  So I uninstalled it, deleted the files from it's folder and then reinstalled the plugin.  Now, when it runs, it doesn't report any issues and the drive listing no longer lists any read (or write) errors, but I'm sure they still exist.   I had previously ran an extended SMART test on disk5 and it the same errors as those noted in the earlier FCP plugin checks. 

 

To prevent further reads and writes to the drives while I diagnose/fix any issues (not sure fix is possible), I stopped the array, removed disk5 from the array share settings, and then restarted the array.  I then tried to run the "reiserfsck --check /dev/sdl" command to check the drive, but I get the following:

~# reiserfsck --check /dev/sdl 
reiserfsck 3.6.27

Will read-only check consistency of the filesystem on /dev/sdl
Will put log info to 'stdout'

Do you want to run this program?[N/Yes] (note need to type Yes if you do):Yes
Failed to open the device '/dev/sdl': Unknown code er3k 127

 

Is this a result of the drive being excluded from the array?  My drives are all formatted as btrfs; I don't recall exactly why I chose this format when I first started the server! 

 

I guess I'm looking for some help in how to check the drive, possibly fixing it, and/or continue using it (maybe that isn't wise?!?!).  I've attached my diagnostics for information/reference.

 

Thanks,

Armand

 

tower-diagnostics-20250114-0623.zip

2 hours ago, ATEglauer said:

# reiserfsck --check /dev/sdl 

This is not the correct command, use the GUI instead, but the disk appears to me mounting, though it may be failing I would recommend running an extended SMART test on that disk

Thanks, JorgeB.  I'm not sure how to check the disk in the GUI.  Clicking on disk5 just refers me to Disk Settings, but I think that only shows me for all the disks and I don't see anything to check the disk other than SMART settings.

http://i.imgur.com/alCooUtm.png

 

I ran an extended SMART test before I moved the data off of it.  The extended SMART report is attached.  I don't know what SCRUB is, but it doesn't indicate any errors.  Should I run that after checking the 'repair corrupted blocks'?  Or should I run another extended SMART check on the disk now that it is empty?

WDC_WD40EFRX-68WT0N0_WD-WCC4E1TH9PTL-20250111-0757.txt

Edited January 14, 20251 yr by ATEglauer

13 hours ago, ATEglauer said:

I'm not sure how to check the disk in the GUI. 

I didn't check before, but the disks are btrfs formatted, not reiserfs, so you just need to run a scrub.

I ran the scrub yesterday; it finished in no time with no errors, likely because there is nothing on the drive??  I checked the repair corrupted blocks before I ran it.

  

 

 

 

I also just re-ran it, with the same results.  I started another extended SMART test, but I received an unexpected 'read failure' in a minute or two.  

 

 

 

 

I've restarted the extended self-test and it seems to be running ... results probably in about 8 hours or so, based on the duration when i ran it before. 

 

The read errors first showed up when I was doing a parity check (which started on Jan. 6 and ran through the nights until Jan 11.  After that, I moved the data from the disk to another disk and the read errors increased.  The disk log information for disk5 is now looking this this (I'm guessing this might be included in the diagnostics, but I couldn't quickly find it):

 

 

 

I'm not sure, but do things appear a little better since yesterday??  Anyway, I'll post the result of the latest SMART self-test when it completes and probably include another diagnostics log at the same time.

 

Armand

Edited January 15, 20251 yr by ATEglauer

26 minutes ago, ATEglauer said:

I ran the scrub yesterday; it finished in no time with no errors, likely because there is nothing on the drive??

Yes, a scrub will only check the actual data, to confirm there aren't any corruptions, I thought that was what you wanted, if you want to check the disk instead, run an extended SMART test.

Extended SMART test seems to have passed, it indicates no errors logged, yet there are 2 more errors (for a total of eight) in the attached SMART report.  I'm really confused, but also guessing it's best not to use this disk???   

 

http://i.imgur.com/qEmEavtm.png

 

*Edit:

I've also attached the latest diagnostics file.

 

 

 

 

WDC_WD40EFRX-68WT0N0_WD-WCC4E1TH9PTL-20250115-2231.txt

tower-diagnostics-20250115-2246.zip

Edited January 16, 20251 yr by ATEglauer

The last test passed, so the disk is OK for now, but it failed the previous test, and likely there will be more errors in the near future, you can keep monitoring for now, especially these attributes:

 

ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
  1 Raw_Read_Error_Rate     POSR-K   200   200   051    -    189
200 Multi_Zone_Error_Rate   ---R--   200   200   000    -    54

 

If they keep increasing or you get more read errors, I would replace that disk.

Thanks very much for your time on this, JorgeB!

 

I will keep monitoring it, but will look at a replacement drive in the near future. 

 

Thanks again, 

Armand

9 hours ago, JorgeB said:

keep monitoring for now, especially these attributes:

 

ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
  1 Raw_Read_Error_Rate     POSR-K   200   200   051    -    189
200 Multi_Zone_Error_Rate   ---R--   200   200   000    -    54

 

If they keep increasing or you get more read errors, I would replace that disk.

 

Unraid monitors some important SMART attributes, and will notify you if they increase. Do you have Notifications setup to alert you by email or other agent as soon as a problem is detected?

 

Those particular SMART attributes are not monitored by default. You should click on each of your WD disks and add 1, 200 to the monitored attributes.

On 1/14/2025 at 9:20 AM, ATEglauer said:

I moved all of the data to drive9 (sde) in the array. 

For future reference, perhaps when a disk is worse than this case.

 

Moving data to other drives in the array is not really a good strategy when a disk is having problems. Moving data actually does additional writing  (deletes) on the problem disk. Usually just replacing the disk and letting it rebuild would be safer.

 

In any case, you must always have another copy of anything important and irreplaceable. Parity is not a substitute for backups. If you feel you need to get data from a failing disk, copy (not move) it off the array. That way no disk in the array is changed, including the problem disk.

22 minutes ago, trurl said:

 

Unraid monitors some important SMART attributes, and will notify you if they increase. Do you have Notifications setup to alert you by email or other agent as soon as a problem is detected?

 

Those particular SMART attributes are not monitored by default. You should click on each of your WD disks and add 1, 200 to the monitored attributes.

 

Hi trurl,

 

I didn't have notifications set up except in the browser, but I have now set up notifications in Discord as well.  As for the SMART settings, can you confirm this is what you meant?

 

 

Right after doing that, I've received a few other warnings.  They might be useful if I had a clue what they meant... guess I need to learn a lot more about SMART errors.

 

 

11 minutes ago, trurl said:

For future reference, perhaps when a disk is worse than this case.

 

Moving data to other drives in the array is not really a good strategy when a disk is having problems. Moving data actually does additional writing  (deletes) on the problem disk. Usually just replacing the disk and letting it rebuild would be safer.

 

In any case, you must always have another copy of anything important and irreplaceable. Parity is not a substitute for backups. If you feel you need to get data from a failing disk, copy (not move) it off the array. That way no disk in the array is changed, including the problem disk.

 

OK, that makes sense of not moving them off the disk.  I was thinking if I moved the data, then parity would be updated, but I suppose it would be with 'copy' as well.

 

 

 

8 minutes ago, ATEglauer said:

can you confirm this is what you meant?

yes

9 minutes ago, ATEglauer said:

received a few other warnings

Click on the SMART warning ( 👎 ) for the disk on the Dashboard page and Acknowledge. It will warn again if it increases.

10 minutes ago, ATEglauer said:

learn a lot more about SMART errors.

Anything that continues to increase should be investigated.

 

12 minutes ago, ATEglauer said:

if I moved the data, then parity would be updated

All write operations in the array update parity at the same time. Write operations include write, delete, move (which does a delete after copy), even format.

 

14 minutes ago, ATEglauer said:

I suppose it would be with 'copy' as well.

Copying somewhere off the array, which is what I recommended, would not update parity. It would leave all array disks including parity just as they were, since only reading would be done from the disk. Copy could even be from a disabled disk emulated by parity, in which case, parity and all other disks would be read instead and the data obtained from the parity calculation.

2 hours ago, trurl said:

 

Those particular SMART attributes are not monitored by default. You should click on each of your WD disks and add 1, 200 to the monitored attributes.

 

Are WD disks known to have more read issues or should I change the settings for the ST drives as well?

 

 

Those attributes have different meanings for other brands, so not really a simple way to monitor them.