I started looking at my system logs during and after a monthly parity check.  These are popping up for some of the drives as it does the check.  SMART reports come out clean for all the drives.  Parity check come out error free.  No performance issues.  I've got several dockers that work flawlessly.

 

I've got ten hard drives connected to two SATA cards and the motherboard.  Errors happen to any of the SATA ports.  I doubt the cards and the motherboard all are defective given checks turn out fine each time.

 

My extensive Google research shows that it's possibly:

 

- the way unRaid handles spinning up the drives from sleep (or that it doesn't spin them up fast enough or in the correct way)

 

- it's a hardware issue with cables (I don't think this is my case.  I've swapped cables and all tests and parity check come out clean consistently)

 

- power supply issue (new high quality power supply @ 650W should handle the 8 mechanical drives and the 2 SSDs fine, also spinning up all the drives at boot or for parity checks at the same time doesn't cause any loggable issues.

 

Jan 16 19:21:54 XORUNRAID kernel: ata27.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen

Jan 16 19:21:54 XORUNRAID kernel: ata27.00: failed command: READ DMA EXT

Jan 16 19:21:54 XORUNRAID kernel: ata27.00: cmd 25/00:c0:a0:69:71/00:00:49:00:00/e0 tag 16 dma 98304 in

Jan 16 19:21:54 XORUNRAID kernel:        res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)

Jan 16 19:21:54 XORUNRAID kernel: ata27.00: status: { DRDY }

Jan 16 19:21:54 XORUNRAID kernel: ata27: hard resetting link

Jan 16 19:21:54 XORUNRAID kernel: ata27: SATA link up 1.5 Gbps (SStatus 113 SControl 310)

Jan 16 19:21:54 XORUNRAID kernel: ata27.00: configured for UDMA/133

Jan 16 19:21:54 XORUNRAID kernel: ata27: EH complete

Jan 16 19:22:37 XORUNRAID kernel: ata27.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen

Jan 16 19:22:37 XORUNRAID kernel: ata27.00: failed command: READ DMA EXT

Jan 16 19:22:37 XORUNRAID kernel: ata27.00: cmd 25/00:c0:a0:8f:90/00:00:49:00:00/e0 tag 19 dma 98304 in

Jan 16 19:22:37 XORUNRAID kernel:        res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)

Jan 16 19:22:37 XORUNRAID kernel: ata27.00: status: { DRDY }

Jan 16 19:22:37 XORUNRAID kernel: ata27: hard resetting link

Jan 16 19:22:37 XORUNRAID kernel: ata27: SATA link up 1.5 Gbps (SStatus 113 SControl 310)

Jan 16 19:22:37 XORUNRAID kernel: ata27.00: configured for UDMA/133

Jan 16 19:22:37 XORUNRAID kernel: ata27.00: device reported invalid CHS sector 0

Jan 16 19:22:37 XORUNRAID kernel: ata27: EH complete

 

Is this something I can ignore for the time being or do I need to customize something to be a little more forgiving of spin up time of the drives?

It looks like a cable problem to me, but that's not the only potential cause. A new high quality power supply is less likely to be a problem if it has a single (and therefore high current) +12 volt rail. What is the make and model of yours?

 

Read this and then post your diagnostics zip.

850W single 12v rail.  Same one linked below.

 

https://www.amazon.com/gp/product/B00IKDETOC/ref=oh_aui_detailpage_o01_s02?ie=UTF8&psc=1

 

xorunraid-diagnostics-20170118-1111.zip

The errors are on 3 disks, disk 5, 6 and 7, all share the second Asmedia controller, what's the controller model? Asmedia are usually 2 port controllers.

 

 

https://www.amazon.com/gp/product/B0177GBY0Y/ref=oh_aui_detailpage_o01_s01?ie=UTF8&psc=1

 

I have two of these and also using the motherboard SATA plugs.  Thought I'd split the drive between two PCI interfaces dsa and have room to expand later.

The controller uses a port multiplier (or to be exact, it uses 2), I've seen issues before on Linux with the Asmedia controller and a port multiplier, however the other one seems fine, so this is what I would do, swap both Asmedia controllers and see if the issues follows the controller or stays with the slot/disks/cables.