Is there any way ensures that vm(s) deployed can recover if the compute-node/vm is booted/restart

asked 2018-10-25 23:19:19 -0500

I am fairly new to openstack and facing an issue regarding recovery of vms. Following is my query –

Is there any way ensures that vm(s) deployed can recover if the compute-node/vm is booted/restart. As our target vm(s) uses compute-hosts under default availability zone (nova). Hence, we are looking if there is any way we can manage ha policy or resume the host's status-quo post reboot or restart at the following level -

i) compute host or ii) availability zone or iii) host aggregate

Please note: We already have 2 options but none of them are viable as none of them restrict it to Availability Zone, Host Aggregate or Compute Host level. i) resume_guests_state_on_host_boot as true in nova.conf : has impact on all the VMs deployed under that compute node ii) making metadata ‘ha:policy=managed_on_host’ on a per-vm basis : about 200 vms to be deployed

edit retag flag offensive close merge delete