First time dealing with Disk in error state

[Hinatanko]

Hi,

 

Software

unRAID Version 6.5.3

 

Hardware

Supermicro X10SRi-F

Intel Xeon E51650 v4 @ 3.6ghz

64GB ECC Ram

 

Array

18x 8TB NAS HDDs (2 of these are parity)

2x 4TB Desktop HDDs

2x 512GB Samsung 850 Pro Cache Pool

 

 

Issue

 

After using unRAID for almost 2 years I got my first notification about a drive throwing an error. The disk was put offline and after doing some research I saw some threads mentioning turning off dockers and not writing to the disks. I have done this already. Attached are the syslogs. The drive that was taken offline is disk9.

 

The server uses a Norco 24 bay case. I am using 20 drives currently (18x 8tb and 2x 4tb) 2 of the 8tb are used as parity.

 

I recently ran extended smart reports about 2 weeks ago and did not get any notifications about issues. I checked my disks again and found that another drive had some 199 errors as a warning.

 

Could someone take a look at my logs and let me know if I just need to replace the drive or is their more steps I need to take to fix this issue? Is their any other drives that might look like they need replaced?

 

Thank you

nas48-diagnostics-20180807-2224.zip

[JorgeB]

Just do a standard replacement, didn't look at all the other SMART reports but as long as you have notifications enable you'll get warned of any SMART issues.

[itimpi]

I assume the problem disk was disk9 which is showing 256 Pending Sectors?   If so it is always possible that rewriting the sectors might fix the problem and the drive could continue to be used.   Ideally you should replace the disk, and then test the old disk when the array is back in a good shape and you are back in a protected state.  If that is impractical you could try rebuilding to the same disk and see if that causes the Pending sectors count to go back to zero.

 

Disk2 is also showing a large value for UDMA_CRC_Error_Count which might indicate a possible cabling problem for that drive if it keeps on increasing (that attribute is never reset so this could indicate an old problem).  While not fatal such errors will cause retries on accessing the disk and thus adversely affect performance.