I came across this error during a parity check:

Nov 15 15:53:21 Tower kernel: ata12.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen (Errors)

Nov 15 15:53:21 Tower kernel: ata12.00: failed command: SMART (Minor Issues)

Nov 15 15:53:21 Tower kernel: ata12.00: cmd b0/d1:01:01:4f:c2/00:00:00:00:00/00 tag 0 pio 512 in (Drive related)

Nov 15 15:53:21 Tower kernel:          res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout) (Errors)

Nov 15 15:53:21 Tower kernel: ata12.00: status: { DRDY } (Drive related)

Nov 15 15:53:21 Tower kernel: ata12: hard resetting link (Minor Issues)

Nov 15 15:53:22 Tower kernel: ata12: SATA link up 3.0 Gbps (SStatus 123 SControl 320) (Drive related)

Nov 15 15:53:22 Tower kernel: ata12.00: configured for UDMA/133 (Drive related)

Nov 15 15:53:22 Tower kernel: ata12: EH complete (Drive related)

 

Probably bad connection, but upon further inspection I notice that this only happens when the web interface is trying to query the SMART report during parity check, e.g. when clicking Refresh on Main to refresh the parity check progress, or unmenu's MyMain which refreshes the progress automatically.

If I don't click the Refresh button, and close the unmenu's MyMain window, I don't get such errors.

 

Might be worth noting that the error doesn't occur during normal operation, such as reading or copying files from/to the affected disks.

 

Also, it seems to only happen the the 13th ports onwards (in my case, disk #13 and #14). The SATA ports are fine, because I took it out, installed and tested it on another system.

 

Another thing that I notice:

- If I use one 8-port controller, the error occurs for half of the ports (4 ports) -and never for the other half, i.e. always the same half.

- If I use two 4-port controllers, the error occurs for the 4 ports at the 'second controller', - and never for the 'first controller'.

 

Check for BIOS and SATA card firmware updates.

Couldn't find update to the card's firmware.

 

but anyway my supermicro card just arrived, and now at early stage of parity check.

So far I don't encounter the condition that I described in the original post, knock on wood.

 

also, as a bonus, the parity check's speed is faster.