I have a cluster of Minisforum BD785i machines with two Intel P4600 U.2 NVMe each and two Crucial 2TB M.2.
I got them working okay-ish, but the logs are showing regular timeout on the NVMe drive’s. It’s mostly the Intel disks, but the Crucial drives have some errors as well.
Initially I thought this was just related to the Intel disks which are refurbished (5% wear), but seeing the Crucial drives also have time time-out has me confuzzled.
It seems there is some issue with the interrupts and messages from the drives being sent but maybe not being received by the chipset/Linux. I’ve turned off SR-IOV and IOMMU, but that’s done nothing either.
They’re all older BIOS versions (1.07, 1.09 Vs 1.15) because I am paranoid about doing an update (I already killed one Minisforum motherboard this way). So it could be AGESA needs updating?
Can anyone suggest something I might be missing?
I think it’s causing some stalling on my VMs rather than just error messages.