Parity Check Speed Crawling After Unclean Shutdown

[cpthook]

Hello forum...  My server failed to shutdown clean after attempting from the web gui "powerdown" button and became hung up.  The last terminal log info showed unraid forcing shutdown after graceful shutdown failed but my machine did not shut off.  I then performed a hard restart from within a IPMI terminal session and unraid came back online.  A parity check began once I started the array but the speed was only between 14.5 - 40MB/s.  My parity speeds are normally in the 75-80MB/s range so it seems they have decreased by less than half.  I powered off the server again to verify ALL sata connections were firm and re-seated them.  SMART test on all drives returned no errors.  I also ran a diskspeed benchmark test and found disk 5 benchmark results not looking like the rest. 

 

Please advise on what may have caused my parity speeds to plummet.  I shared the diskspeed benchmark results below along with diagnostics.  

 

Thanks, in advance for any assistance provided!!!      

 

 

 

 

tower-diagnostics-20190308-1747.zip

Edited March 9, 2019 by cpthook
Spelling errors

[tr0910]

Parity runs as fast as your slowest disk. It sure looks like disk 5 is behaving strangely. In a situation like this I like to have a spare tested and pre cleared that I can confidently have ready to replace the problem disk.

However you may find this is a transitory problem. Good to let it finish the parity and the test that disk out with a few more reboots and change its data cable connection to make sure it's disk not controller issues. Let it run for a bit, but watch it closely.

Sent from my chisel, carved into granite

[JorgeB]

5 hours ago, cpthook said:

I also ran a diskspeed benchmark test and found disk 5 benchmark results not looking like the rest. 

That disk is showing some warnings:

Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
  1 Raw_Read_Error_Rate     POSR-K   200   200   051    -    11035
 200 Multi_Zone_Error_Rate   ---R--   200   164   000    -    57

These usually are caused by slow sectors, and are consistent with bad performance in some zones, and they tend do progress to bad sectors sooner or later.

[cpthook]

10 hours ago, tr0910 said:

Parity runs as fast as your slowest disk. It sure looks like disk 5 is behaving strangely. In a situation like this I like to have a spare tested and pre cleared that I can confidently have ready to replace the problem disk.


 

Thanks million for the response!!! Luckily for me, I do have a drive cleared and ready to go.

 

7 hours ago, johnnie.black said:

That disk is showing some warnings:

Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
  1 Raw_Read_Error_Rate     POSR-K   200   200   051    -    11035
 200 Multi_Zone_Error_Rate   ---R--   200   164   000    -    57

These usually are caused by slow sectors, and are consistent with bad performance in some zones, and they tend do progress to bad sectors sooner or later.

 

Thanks for the re-assurance that drive 5 looks bad.  I kinda figured so based on the benchmark tests, however, I wasn't 100%.  I always like to consult the community first for a 2nd opinion before making any adjustments or repairs.  Good news is that once I replaced drive 5 with a new cleared disk, parity speeds are back to normal.  Well, at least the "parity rebuild" process speed is where is should be, but parity sync speed should be the same or close.

Edited March 9, 2019 by cpthook
Spelling errors