Problems with LSI 9300-16i?

[mono_void]

Hello everyone! I come here when Reddit doesn’t cut it.

I recently rebuilt my TrueNAS Scale server to get away from an old power-hungry Xeon. I finally decided to get a proper HBA, and I got an LSI 9300-16i from eBay.

After I spent a few hours doing the build and install, I noticed I could not get past grub. I started taking things out of the system, and luckily, I started with the HBA. As soon as I took it out, it went straight to the TrueNas installation. After I finished, I put the HBA back in, and logs immediately flooded the console (see picture). The errors happen with or without drives plugged into the HBA. If I ignore the error, I can see each drive from the GUI or within the shell.

I made a post to Reddit about it and got some advice; I tried everything, which included:

1. Flashed firmware that iXSystems worked with Broadcom for. You can find this on the iXSystems forums.
2. Flashing the latest firmware that I got from the Broadcom website.
3. Editing grub to include these parameters:

pci=nomsi
pci=noaer
pci=nommconf
pcie_aspm=off

I am not sure if I did it correctly. But I would save it one at a time and reboot. Obviously, I still got the error.
4. I also tried to switch on and off a ton of different things in the BIOS, such as power management settings such as ASPM, and more.
5. I can see the HBA in the BIOS; the only option I have to change is Legacy (on or off). I tried both, no dice.

I am really no expert at any of this, and I am also dyslexic as hell. Ha! If anyone has got some input on what I should do here, please help! Give me a little time to respond, I am very busy with school and work. Thanks for the help y’all!!!

OS: TrueNAS-SCALE-23.10.2
Specs:

• MOB: GIGABYTE Z690 GAMING X
• CPU: Intel 12th Gen i5-12600K
• RAM: 64GB DDR4
• PSU: SEASONIC PRIME 750W 80+T
• NIC: Intel X520-DA2 10Gb

[EniGmA1987]

Im not expert, but I see physical layer errors so I would make sure it is seated right in the pcie slot. It could also be that the used HBA you got has hardware problems and just isnt working right.

[strh]

Stab in the dark… power draw? I think the card is 27W nominal.

[qqk]

Two things to try:

1. That card is two 8-port SAS3008 chips glued together with high power requirements (27W] and consequently high heat dissipation. Notice the huge heatsink with no fan? It’s designed for server cases with high front-to-back airflow. Try putting a spare 80-120mm fan pointed towards the card to see if it makes a difference with the errors.

2. In your BIOS, manually reduce the cards PCIe link speed to Gen 2 and even Gen 1 if needed. Does that stop or reduce the errors?

Otherwise, you may have a faulty card on your hands.

[mrpasc]

You hopefully connected that additional power connector (6 Pin) to a PCIE power cable at your PSU?

[mono_void]

I see a 6-pin out that looks like it can take extra power from the PSU, but I wasn’t sure if I should try that. I am confused about the power requirements here. Doesn’t a consumer MOB give 75wats to a 16x slot? Shouldn’t that be enough?

[bambinone]

Yes, but it’s a x8 card, and an x8 slot is only guaranteed to provide 25W. In other words, they had to engineer the power delivery circuitry under the assumption that it wouldn’t be able to draw the full 27W from the slot.

[mono_void]

ah. that makes sense. thank you!