How to trigger instance to be "crash" status to test the on_crash action?

asked 2015-12-29 20:05:17 -0600

xiangfei-ma gravatar image

I am running Kilo on RHEL7 with libvirt 1.2.8 and qemu 1.5.3,
the xml shows that default on_crash action of instance is destroy like below:


My customer asked me to change the on_crash action from destroy to restart, and I did it.
But now I don't know how to trigger instance to be "crash" status to verify the restart process is working fine.
Can someone tell the sample steps to trigger instance to be "crash" status?
Or, tell the mechanism that how dose qemu/libvirt detect and judge the instance is "crashed"?

Much appreciated.

edit retag flag offensive close merge delete