HDDs getting randomly disabled.

[negurici]

Hello.

I have an issue that started a few weeks ago. Everything was running fine, disks were reported as healthy, until one morning when I saw that Disk 5 was reported as disabled. Tried to reseed the cables, removed the disk from the array, started it in maintenance mode, stopped the array and re-added Disk 5 to the array and started it normally and the parity rebuild was started. Everything was ok for 2-3 days after which it was disabled again. This time I changed the cables, did the steps to add the disk back to the array, and it all went well for a couple of days. Afterwards, the same thing happened, only that Disk 4 was the one that kept being disabled. Now the same happens with Disk 2. The weird part is that all the drives appear to be healthy if I do a S.M.A.R.T. check. I also did a BIOS update, but that didn't seem to change anything.

I have noticed that for Disk2, Disk3, Disk4 and Disk5 sometimes don't report the temperature in Unraid, even if they are not spined down.

Another thing that they have in common is that all 4 drives are connected directly to the motherboard (ASUS PRIME B450M-A II). The other 4 drives are connected to a PCI-E card, and they don't have any issues, but that being said, they are also the newer ones.

My hunch would be that there's an issue with the SATA controller on the motherboard, but I don't know how I could investigate if that's the issue.

I could try to change the motherboard with an older one I have (GIGABYTE B450 I AORUS PRO WIFI) but for that one I wouldn't have enough space to connect both the video card and the PCI-E SATA controller, but I don't know what that will imply for the Unraid installation.

tower-diagnostics-20220706-1318.zip

[JorgeB]

Looks like the typical Ryzen onboard controller issue, it's quite common with some Ryzen servers, usually mostly under load, look for a BIOS update but best option might be to use an add-on controller.

[negurici]

Thank you! Adding an add-on controller solved the issue.