X470D4U - SATA fails under load

Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic

   I have the aformentioned board, running BIOS 3.30.

I have two WD Ultrastar 14TB drives (ZFS mirrored) in two of the sata ports, but when I perform "zpool scrub", which performs heavy disk I/O, after some period of time the main machine will power-cycle-reset.  I get no kernel information save a sata packet timeout (I'm running OmniOSce, an illumos kernel).

I understand it could be the drives, and I'll look into that separately.  I'm asking here about:

- Has anyone seen SATA failures like this at all on this board?

- Does the 3.50 BIOS (which I'll be flashing later) have known SATA  improvements?  I saw the one person complain about HW-RAID with 3.50, but I don't trust HW-RAID of any flavor, hence ZFS.

Thanks.

Author	Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic
danmcd Members Profile Send Private Message Find Members Posts Add to Buddy List Newbie Joined: 22 Feb 2021 Status: Offline Points: 5	Post Options Post Reply Quote danmcd Report Post Thanks(0) Quote Reply Topic: X470D4U - SATA fails under load Posted: 22 Feb 2021 at 12:36am
	I have the aformentioned board, running BIOS 3.30. I have two WD Ultrastar 14TB drives (ZFS mirrored) in two of the sata ports, but when I perform "zpool scrub", which performs heavy disk I/O, after some period of time the main machine will power-cycle-reset. I get no kernel information save a sata packet timeout (I'm running OmniOSce, an illumos kernel). I understand it could be the drives, and I'll look into that separately. I'm asking here about: - Has anyone seen SATA failures like this at all on this board? - Does the 3.50 BIOS (which I'll be flashing later) have known SATA improvements? I saw the one person complain about HW-RAID with 3.50, but I don't trust HW-RAID of any flavor, hence ZFS. Thanks.

danmcd Members Profile Send Private Message Find Members Posts Add to Buddy List Newbie Joined: 22 Feb 2021 Status: Offline Points: 5	Post Options Post Reply Quote danmcd Report Post Thanks(0) Quote Reply Posted: 22 Feb 2021 at 12:52pm
	One other thing. It might not be related because its timing was nowhere near the spontaneous power-cycle, but I saw this in the IPMI event log: 42 \| 02/21/2021 \| 00:30:18 \| OEM record c1 \| 000000 \| 06020a010108 Not sure what the OEM code bytes (06020a010108) mean, but ASRock Rack sure does, I suspect. Clues welcome! :)

danmcd Members Profile Send Private Message Find Members Posts Add to Buddy List Newbie Joined: 22 Feb 2021 Status: Offline Points: 5	Post Options Post Reply Quote danmcd Report Post Thanks(0) Quote Reply Posted: 22 Feb 2021 at 10:39pm
	Update: Same misbehavior with BIOS 3.50.