SOLVED TrueNAS Core PCIe NVMe Card Issues

[adityaharsh]

I've an Asus HYPER M.2 X16 Gen 4 card with 4x Samsung 980 Pro 1TB, this card was working fine inside TrueNAS 12 U8, but when I upgraded my hardware and moved to TrueNAS 13 U1, this card started giving this error screen (attached image below)

All SSDs work fine as they normally should when I connect them to the Motherboard, but my MoBo only has 2 M.2 Slots. This card is working fine, I tested it in Windows 11, Ubuntu 22.04 LTS.

Any suggestions as to where it might go wrong?

My specs:
AMD Epyc 7542
Supermicro H12SSL-I
Micron 64GB ECC RDIMM

 

[mav@]

These errors are not from the NVMes, but from LSI HBA. Somehow its firmware does not cooperate with the driver.

 

[adityaharsh]

These errors are not from the NVMes, but from LSI HBA. Somehow its firmware does not cooperate with the driver.

But I've been using this HBA for quite some time (almost 8-9 months) and it has never given me any issues. Also this is not permanent but very random.
Sometimes, it goes away with hard reset and sometimes it takes multiple tries. I never had this kind of issues in 12 U8.1 but after 13 U1 release I was convinced to update. Any settings or kernel options I can change to make this smoother.

 

[mav@]

I can't guess what exactly goes on there from just that screenshot and how can it be related to the upgrade. Make sure you are not overheating the HBA.

 

[adityaharsh]

I can't guess what exactly goes on there from just that screenshot and how can it be related to the upgrade. Make sure you are not overheating the HBA.

Any other information you require that can be helpful in solving this?

 

[Elliot Dierksen]

I agree with @mav@ that overheating is a likely cause. Those cards do run hot, and overheating makes them act very strange.