**RESOLVED** Parity check finding 7000 errors

[Fiservedpi]

Is this a sign of something bad?

Edited January 19, 2020 by Fiservedpi

[JorgeB]

Any unclean shutdown since last check? A few sync errors are expected after an unclean shutdown.

[Fiservedpi]

No, here's the smart report if I'm reading correctly all my hard drives are in prefail? All drives are less than 4 months old I did relocate the server 18 days ago

tower-diagnostics-20200118-1143.zip

[itimpi]

14 minutes ago, Fiservedpi said:

No, here's the smart report if I'm reading correctly all my hard drives are in prefail? All drives are less than 4 months old I did relocate the server 18 days ago

tower-diagnostics-20200118-1143.zip 157.97 kB · 0 downloads

The drives look fine as far as I can see.   What made you think they were in pre-fail?

[Fiservedpi]

i was never good at reading at the smart reports but i thought P = Prefail 

1 Raw_Read_Error_Rate     POSR-K   200   200   051    -    0
  3 Spin_Up_Time            POS--K   167   167   021    -    6608
  4 Start_Stop_Count        -O--CK   100   100   000    -    385
  5 Reallocated_Sector_Ct   PO--CK   200   200   140    -    0


                            ||||||_ K auto-keep
                            |||||__ C event count
                            ||||___ R error rate
                            |||____ S speed/performance
                            ||_____ O updated online
                            |______ P prefailure warning

Which sections should i keep an eye on the smart report?

Edited January 18, 2020 by Fiservedpi

[itimpi]

9 minutes ago, Fiservedpi said:

i was never good at reading at the smart reports but i thought P = Prefail 

1 Raw_Read_Error_Rate     POSR-K   200   200   051    -    0
  3 Spin_Up_Time            POS--K   167   167   021    -    6608
  4 Start_Stop_Count        -O--CK   100   100   000    -    385
  5 Reallocated_Sector_Ct   PO--CK   200   200   140    -    0


                            ||||||_ K auto-keep
                            |||||__ C event count
                            ||||___ R error rate
                            |||____ S speed/performance
                            ||_____ O updated online
                            |______ P prefailure warning

Which sections should i keep an eye on the smart report?

I would normally just keep an eye on the ones that Unraid monitors by default (as set under Settings->Disk Settings).   You then get notifications (assuming notifications are enabled) when these change.

 

The only exception might be if you have run a SMART extended test when you want to explicitly go in and see if it passed/failed independently of the SMART attributes.

[Fiservedpi]

Ok thanks which attributes are indicative of a failing drive

[PeteAsking]

The seagate has a high seek error rate but I dont really understand how to interpret that, maybe someone else knows if thats normal or bad:

 

Seek_Error_Rate         POSR--   085   060   045    -    352968323

[Fiservedpi]

Why can't they just write it in English lol

"yo disk f#$@ed buy a new one"

 

Plug-in idea? Haha 

Edited January 18, 2020 by Fiservedpi

[itimpi]

3 minutes ago, Fiservedpi said:

Ok thanks which attributes are indicative of a failing drive

The ones that you want to monitor especially are the Reallocated Sectors and Pending Sectors.   reallocated Sectors are not necessarily a problem as long as the number stays small and stable, but a large number or a continually increasing number are a bad sign.    Pending Sectors can be intermittent, but they are always working as they indicate there are sectors that may not be read correctly if another drive fails and a rebuild is necessary.   However they not infrequently get cleared when the sector is next written so again not necessarily fatal if they only happen occasionally.    The CRC count is one to monitor as it indicates a connection problem (typically cabling) rather than a disk problem, but they can still cause a read or write to fail causing Unraid to disable the drive.   CRC count never gets reset - it only increases so keeping it stable is what you aim for.

[itimpi]

5 minutes ago, PeteAsking said:

The seagate has a high seek error rate but I dont really understand how to interpret that, maybe someone else knows if thats normal or bad:

 

Seek_Error_Rate         POSR--   085   060   045    -    352968323

I believe that for Seagate that is a bit-dependent field (so not really easy to interpret) rather than a raw count.

[PeteAsking]

1 minute ago, itimpi said:

I believe that for Seagate that is a bit-dependent field (so not really easy to interpret) rather than a raw count.

Thank you. 

[Fiservedpi]

Just upgraded to 6.8.1 re-ran parity check found 0 errors so guess it was a false alarm thanks for the info everyone