So i've seen these errors before (not frequent at all) on my server but they've "fixed themselves" so i didn't bother investigating further. However now it's becoming an actual problem since it's never ending and ended up corrupting stuff once (guess i should've not ignored it in the first place huh).

 

Jul 1 22:19:51 Henkraid kernel: ata9.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
Jul 1 22:19:51 Henkraid kernel: ata9.00: failed command: WRITE DMA EXT
Jul 1 22:19:51 Henkraid kernel: ata9.00: cmd 35/00:08:58:e1:ea/00:00:e8:00:00/e0 tag 17 dma 4096 out
Jul 1 22:19:51 Henkraid kernel: res 40/00:01:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
Jul 1 22:19:51 Henkraid kernel: ata9.00: status: { DRDY }
Jul 1 22:19:51 Henkraid kernel: ata9: hard resetting link
Jul 1 22:19:51 Henkraid kernel: ata9: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Jul 1 22:19:51 Henkraid kernel: ata9.00: configured for UDMA/133
Jul 1 22:19:51 Henkraid kernel: ata9: EH complete

 

Worth to mention is it doesn't happen frequently at all if i just let the system run normally, but a few days ago i removed 3x1tb drives and added 2x10tb drives and started preclearing them. The preclear combined with docker containers running will trigger this to happen, and it won't stop unless stuff start getting corrupted and unRaid decides the SSD(s) are faulty or i reboot the server when the log won't stop filling with it.

I've tried what i can do hardware wise so far, different sata cables, different ports both on motherboard and expansion card, pulled the GPU to make sure there's enough power for the drives, i even precleared the ssd's to see if it reported anything weird but it finished successfully (no idea it that was a good idea).

 

It can handle a parity check + two preclears at the same time but with no docker containers running, so it has something to do with the ssd's i think. I don't believe they are broken though and it's starting to feel like a hardware error, but what do i know... Will try with a different PSU as soon as i can find the right cables i need for my spare one.

 

If anyone have any idea about this please do tell, after i've tried another PSU i don't know what to do. Thanks.

henkraid-diagnostics-20190701-2109.zip

Seldom found report for SATA error with onboard FCH, you have 2 drive have this problem, but both error cause was different.

 

As you have 2 same SSD, I would suggest you swap their SATA port which connect at mainboard side, if problem still on same SSD, that should be SSD problem rather than PSU, cable .... ( I also notice you mention, seems problem only occur when docker running, weird )

 

[9:0:0:0]    disk    ATA      WDC WD40EFRX-68N 0A82  /dev/sdd   /dev/sg3 
[13:0:0:0]   disk    ATA      INTEL SSDSC2KW51 002C  /dev/sde   /dev/sg4 
 

 

Jul  1 21:53:50 Henkraid kernel: ata13.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
Jul  1 21:53:50 Henkraid kernel: ata13.00: failed command: FLUSH CACHE EXT

 

Jul  1 22:19:51 Henkraid kernel: ata9.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen
Jul  1 22:19:51 Henkraid kernel: ata9.00: failed command: WRITE DMA EXT

 

 

It's not always on the same SSD, and this is the first time i saw it on the WD drive. Moved all 6 drives to the motherboard connectors now and only use the expansion card for the two i'm preclearing, and it seems fine right now actually. Might be a problem with that card then, assuming the errors won't come after a while.

The 6 or 10 port Asmedia controller you're using is just a two port SATA controller with two SATA port multipliers, they are known to cause a lot of timeout issues, and the server might be unresponsive during those, they are not recommended for Unraid.

I've got some DeLOCK 10x sata expansion card.

They are available as 6 or 10 port controllers, by various manufactures, but all have the same problems.