Comment 1 for bug 1647021

Hi, Xuchao,

Sometimes when traceback is shown, it doesn't really mean there is a bug... Regard of this, I would say there might be something wrong with your OpenStack.

What VMTP trying to do now is, creating two VMs on the same compute node and same network, and start the nuttcp between this two VM. One thing you could try is, run VMTP with --stop-on-error. Once the error occurs, everything will be remained as is without cleanup. Then you could try to ssh to one of the VM and try to ping and diagnose more on it.

Regards,
Yichen