Disk Errors

[B_Sinn3d]

Hey Everyone,

 

So long story short.  I just built a new dual core Xeon build.  I moved all my drives and memory to the new case.  When I moved the SuperMicro 5 drive cage, I accidentally unplugged fan so my drives started running hot and came home to alarms blaring.  Drives were hitting mid 50's Celsius for a couple of hours intermittently as there were being used.

 

I powered down server, re-attached the fan and powered server back up.  Disk 6 dropped offline.  I assumed I knocked the cable when re-attaching the fan so I stopped the array, removed disk, stopped and added disk back in.  During the rebuild I am now getting reallocated sector count errors on one of my parity (Dual Parity) drives.  Drives 3 and 4 also were showing thousands of errors so I stopped the rebuild.  Not sure where I should go from here.  

 

I am running a extended smart test on parity 1, disk 3, 4 & 6 now.  Attaching diags from earlier.

 

Any help would be appreciated.

debo-server-diagnostics-20180331-2216.zip

[JorgeB]

You never want to use the Marvell controller on those boards, the 4 first four white SATA ports, they are known to drop disks all the time, the reallocated sectors on the parity disk are a disk problem but guess where the other problem disks are connected?

[B_Sinn3d]

Thanks Johnnie, I made an assumption that the marvel controller was the 4 ports away from other 10.  I did read some about issues with Marvell controller but thought the iommu=pt to the go file would help with drives dropping off array. 

 

I moved everything off Marvell controller and added my dell perc 310 to give me the needed sata ports.  It is doing the rebuild of disk 6 now.

 

Hopefully all goes well. I appreciate your assistance!

 

DB