I ran into a similar issue where I’d get freezes with no logs etc. It looks like there’s a known bug in the 5.14.x kernels related to the I/O scheduler.
Here are the bugs:
https://bugzilla.redhat.com/show_bug.cgi?id=2008529
https://bugzilla.kernel.org/show_bug.cgi?id=214503
As the kernel bug suggests, it could be a scheduler bug. So, can you check what your scheduler is currently?
cat /sys/block/*/queue/scheduler
If it’s bfq, that could be causing it. Try changing it to mq-deadline:
echo mq-deadline | sudo tee /sys/block/sd*/queue/scheduler
Change the sd* bit here depending on what your drives are called.