Issue

• Openstack platform shows the VM "running" state on the compute node. Libvirt logs did not show any symptom of the instance rebooting at hypervisor level.
• On the Hypervisor after the application hangs we see:

kernel: ixgbe 0000:0a:00.0 enpX: VF Reset msg received from vf 15  

and:

[ ] Detected Tx Unit Hang
[]   Tx Queue             <0>
[]   TDH, TDT             <0>, <2>
[]   next_to_use          <2>
[]   next_to_clean        <0>
[] tx_buffer_info[next_to_clean]
[ ]   next_to_watch         <ffff880b860b8010>
[ ]   eop_desc->wb.status  <168080>
[]   time_stamp           <fffb97a8>
[]   jiffies              <fffb97aa>
[] Detected Tx Unit Hang
[]   Tx Queue             <0>
[]   TDH, TDT             <0>, <2>
[]   next_to_use          <2>
[]   next_to_clean        <0>
[] tx_buffer_info[next_to_clean]