9300-8i HBA Problem

DonJuanMD · February 25, 2023

I have a 4 year old LSI HBA 9300-8i behaving weird one day. It wont boot to its bios and showed the "fault 05h" message. I posted recently about this and a flash to a new firmware was advised. I havent flashed an HBA card before so I was looking around the forum and the internet and no luck for me and was not able to find the specific methods to get it to work from making a USB boot drive to make the sas3flash to boot into DOS. I then tried to boot it in another machine, still no luck there. A few hours ago I tried to install the card back to the machine and then miraculously it booted. I check the firmware and it is on ver 16.0.0~. So I figured it does not need a flashing to a new version. I turned off the machine so that I can slap back all the drives and then on reboot it failed again and wont boot giving me the fault 05h. I suppose this could be an overheating problem and opened the card and here are my findings. Any thoughts?

leadeater · February 25, 2023

Sounds like corrupted firmware, re-flash it even though it might the the current latest version. Also as a test disconnect all the disks and see if it boot properly. If it failed every time the disks are connected then you could have a faulty disk that is caused this problem.

DonJuanMD · February 25, 2023

4 minutes ago, leadeater said:

Sounds like corrupted firmware, re-flash it even though it might the the current latest version. Also as a test disconnect all the disks and see if it boot properly. If it failed every time the disks are connected then you could have a faulty disk that is caused this problem.

Thank you for the reply. I checked all the drives and no faulty disks. Loaded all the disks to a different adapter. I also noticed while replacing the thermal paste of the HBA card, I see a chip on the periphery of the card. You may refer to the pictures I attached. I dont know if it is really major. Been using the card for almost 4 years.

Jarsky · February 27, 2023

Sounds like either a firmware corruption as @leadeater mentioned, or some sort of hardware fault.

Follow his suggestion first about re-flashing the firmware, though since it booted again once fine I think it could be the latter.

The LSI BGA chip you posted is the ROC (Raid on Chip) which is the ROM and what does all the computation.I think its highly unlikely from a hardware perspective thats where the issue is since you got it to boot again. There are a number of SOIC's on the board though, and I would hazard a guess that perhaps one of them has a dry leg and something isnt switching properly.

The below SOP-8 chip with the 8 legs is what im talking about. You probably have a few of these. You could use a small pair of tweezers and some sort of magnification to just nudge each leg and make sure theyre properly seated and not cracked joints. It could be that one of these just needs solder to be reflowed.