Raidz1 suddenly Degraded/Corrupted

[retroip]

Hello there,
I'v check few similar topics, but none of them was describing my issue:

Facts:
Version: TrueNAS-SCALE-22.02.0.1
CPU: Athlon 200GE
OS: On M2 ssd (64gb)
MB: Gigabite A320m (4x SATA on board)
RAM: 16GB - nonECC
LSI: LSI-SAS2008 (Firmware 20.*) - 8 port (4+4)
Drives: 12x HDD (1x SSD)
Pools: 5-6 (mirrors, direct, 1xraidz1)

I'm running multiple VMs on raidz1 pool with 4xHDD. Suddenly degraded, suddenly not available for import.

Reproduction:
1.Import via TrueNas GUI - i/o error - destroy and recreate pool)
2.All four HDD fully deeply checked (~5h test per HDD) - no issue, no errors in smart, readable entire
3.zpool import (with 4 drives), hangs entire system, need to reboot - kernel panic.
4.After removing 1 drive (which has lost partition layout 93149....) - I have kernel panic with CPU ERROR (screen below)
5.After removed 1 drive and entering: zpool import -o readonly=on I was able to read my data and backup them.

Questions:
1.Could be CPU an issue ? (to weak) ? Should I buy Ryzen 3+ ?
2.CPU and MB are the cheapest, could be this issue ?
3.How could that 1 drive lost partition layout ?

Many thanks for support and any hint.

 

[Redcoat]

Power supply too small for load?

 

[retroip]

Power supply too small for load?

Brand new 700W PS (mid range) = no clicks even during hot swaps (not performing daily, just for tests)