Question LSI 9266-8i randomly dropped offline and can't get it back

[mac_angel]

I have two 9260-8i (one an IBM OEM) and I got a 9266-8i a few months ago to upgrade one. Built a new RAID 5 and been running it since April with no problems. Yesterday I was streaming a video from the RAID, and setting up some downloads for it, which I've done many times in the past. This time the video cut out, and about a minute later (I was logged in remotely) the drive I was streaming from dropped offline. I opened up MSM and the card wasn't there. I opened up the device manager and it was there, no error messages or anything, "the device is working properly". I restarted a couple of times and it came back, RAID drives and all. The family and I were watching a movie, streaming from the RAID and it cut out a few times, just for a few seconds, but then at the end of the movie, it cut out completely. MSM either doesn't see the card, or lists it as 0 (no information. No firmware, drives, nothing). It is still showing up in the device manager as no problems. Tried restarting many times, no luck. Tried swapping the PCIe slot, no luck. The other 9260-8i is still fine, no issues in either slot.
No updates (on purpose) on the computer. Windows 11 Ghost Spectre, so updates are turned off. I didn't do any driver updates or anything.
I haven't tried the WebOS yet. I'm going to check that after lunch. I have the server in the basement in a room that's not really comfortable to sit at a computer for long.

EDIT: I forgot to add, there's also a 60mm fan strapped onto the cards, right on the heatsink, so heat really shouldn't be a problem.

 

[Misgar]

How hot is the heatsink on the chipset? Is it too hot to keep your finger on for more than 3 seconds?

An LSI 3442 card died in one of my desktop PCs in 2022. No problem with similar cards in my HP ML350p Gen8 servers with forced air cooling.

I replaced the controller and retro-fitted 30mm and 40mm fans on all my LSI cards in various desktop PCs.

Overheating may be a issue if you're using server-class cards in computers with low air flow.

Sorry, I've just spotted your statement about a 60mm fan.

 

[mac_angel]

How hot is the heatsink on the chipset? Is it too hot to keep your finger on for more than 3 seconds?

An LSI 3442 card died in one of my desktop PCs in 2022. No problem with similar cards in my HP ML350p Gen8 servers with forced air cooling.

I replaced the controller and retro-fitted 30mm and 40mm fans on all my LSI cards in various desktop PCs.

Overheating may be a issue if you're using server-class cards in computers with low air flow.

Sorry, I've just spotted your statement about a 60mm fan.

yea, I know they get hot AF, so the fan is needed. No problem touching it. I went into the WebOS to check in there and it starts up there no problem. Sees all the drives, all the information, etc. Temperature read at 58°C, which is really good for these cards.
Restarted, went into Windows, and when I opened Explorer, the drive was there, but when I logged into MSM, which took longer than usual, the card was back to showing 0, and the RAID drive was no longer there. Went back into the WebOS, and everything shows up there same as before.
I'm really leaning towards something in Windows got F'd up, though I don't know what. As I said, Ghost Spectre, so no updates. The only thing that 'changed' was last Monday I took out my son's RTX3050 to give back to him, and put in an older GTX750ti. Not a lot of room in the case with all the hard drive cages. But, that was Monday, and has been fine with everything until yesterday afternoon.

 

[Misgar]

I'm guessing your card could be second hand. If so it might have been subjected to high temperatures in a previous system.

I'm running an 8-port LSI card in TrueNAS Core with 8 SAS hard disks.

The 3442's are for backup to external SAS LTO drives.

 

[mac_angel]

I'm guessing your card could be second hand. If so it might have been subjected to high temperatures in a previous system.

I'm running an 8-port LSI card in TrueNAS Core with 8 SAS hard disks.

The 3442's are for backup to external SAS LTO drives.

it's possible, but odd for it to work all this time and now giving throubles. It shows in the device manager, and WebOS.