Still getting this issue on this version 9.3.
e1000e 0000:00:19.0 eth0: Detected Hardware Unit Hang:
e1000e 0000:00:19.0 eth0: Reset adapter unexpectedly
This issue has been present since 9.1, when there will be a fix?
e1000e 0000:00:19.0 eth0: Detected Hardware Unit Hang:
e1000e 0000:00:19.0 eth0: Reset adapter unexpectedly
Nah, using a hypervisor as workaround only creates more fuss, overhead, wastes energy, consumes time and requires additional knowledge to setup. Also it may introduce other problems, as other products aren't bug-free as well ;-). Sounds like shooting a bird with a pumpgun.
Simply create your own udev rule, replace the vendor & device IDs with the ones of your NIC that triggers the corresponding workaround-script (/lib/udev/nic-disable-[gro|tso|aspm]) as posted above and you're set.