Parity "Errors" - SOLVED

[bonustreats]

Hi Everyone,

 

Setup: 1x4TB WD Red (parity), 2x2TB WD Green, 2x2TB WD Red, 64 GB SSD Cache, 2GB Flash.

 

On the Main tab, under the "Array Devices," unRAID is saying that there are 239,435 errors for the parity disk. However, under the Parity History, the last check completed with 0 errors (run on 14AUG) - please see attached image.  Is this something to worry about?  I tried searching the forums, but could only really find info about actual parity errors that come up during a check.  

 

Diagnostics are also attached.

 

Thanks in advance!

radagast-diagnostics-20170816-1803.zip

Edited August 18, 2017 by bonustreats

[FreeMan]

As I understand it, those are read errors on the disk itself, not parity errors - the errors just happen to be reading the parity disk.

 

HOWEVER - wait until one of the experts peruses the diagnostics file and chimes in with a more authoritative answer before doing anything. In the mean time, I would suggest that you suspend any write activity to your array. If you're simply streaming media, you should be OK until then.

[JorgeB]

12 minutes ago, bonustreats said:

Is this something to worry about?

 

Yes, and they've been happening for a month at least, do you have system notifications enable?

 

Start by running an extended SMART test on the parity disk.

[bonustreats]

2 minutes ago, FreeMan said:

As I understand it, those are read errors on the disk itself, not parity errors - the errors just happen to be reading the parity disk.

 

HOWEVER - wait until one of the experts peruses the diagnostics file and chimes in with a more authoritative answer before doing anything. In the mean time, I would suggest that you suspend any write activity to your array. If you're simply streaming media, you should be OK until then.

Gotcha, I was guessing/hoping it's something along those lines.  Write activity is suspended.  

 

Thanks!

[bonustreats]

6 minutes ago, johnnie.black said:

 

Yes, and they've been happening for a month at least, do you have system notifications enable?

 

Start by running an extended SMART test on the parity disk.

I don't think I have them enabled; is this the correct way to enable them?: https://wiki.lime-technology.com/Configuration_Tutorial#Install_UnMENU_for_Screen_and_Email_Notifications

 

Extended SMART test is currently running, but I think it said to allow 2 hours per TB.  I can just download the results and paste them here, correct?  

 

Thanks for responding!

[bonienl]

6 hours ago, bonustreats said:

I don't think I have them enabled; is this the correct way to enable them?: https://wiki.lime-technology.com/Configuration_Tutorial#Install_UnMENU_for_Screen_and_Email_Notifications

 

No, notifications are enabled under Settings -> Notification Settings.

 

[bonustreats]

18 hours ago, johnnie.black said:

 

Yes, and they've been happening for a month at least, do you have system notifications enable?

 

Start by running an extended SMART test on the parity disk.

Here  it is, sorry for the delay!

 

Edit: sorry, uploaded the wrong one.

 

radagast-smart-20170817-1237.zip

Edited August 17, 2017 by bonustreats

[JorgeB]

Quote

Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
# 1  Extended offline    Completed: read failure       10%     26563         1751758408

 

Replace the disk.

[bonustreats]

4 hours ago, johnnie.black said:

 

Replace the disk.

Gotcha.  Do read failures always equal disk replacement?  Is it safe to assume that any time there are errors in that column on the main screen that the drive is bad?

 

Thanks very much!

[JorgeB]

52 minutes ago, bonustreats said:

Is it safe to assume that any time there are errors in that column on the main screen that the drive is bad?

 

No, read errors on unRAID can be caused also by a bad cable/connection, controller, power, etc, read failure on a SMART test on the other hand it's the disk for sure, when it's a few bad sectors some more adventurous folks like to try and preclear the disk, it may last some more time, but in this case the disk failed on some many different sectors that IMO it would be a waste of time and big risk, any read failure during e.g. a rebuild, will result in a corrupt rebuild.

[bonustreats]

2 hours ago, johnnie.black said:

 

No, read errors on unRAID can be caused also by a bad cable/connection, controller, power, etc, read failure on a SMART test on the other hand it's the disk for sure, when it's a few bad sectors some more adventurous folks like to try and preclear the disk, it may last some more time, but in this case the disk failed on some many different sectors that IMO it would be a waste of time and big risk, any read failure during e.g. a rebuild, will result in a corrupt rebuild.

Yeah, I think I'd be leery about trying a preclear again regardless - seems like you're just begging for trouble.  New one should be here tomorrow.  Thanks very much for your time and responses!  I'll mark this thread as "solved."