Nova Instance Stuck

1.Identify host for stuck instance:

nova list --fields host,instance_name,status | grep -i <instance-name-or-id>

2.Attempt to SSH to the host. If the host is inaccessible, it will need to be made accessible (e.g. reboot or bring up interface which is downed. On the host check that the nova-compute service is running and if not, start it:

service nova-compute status
service nova-compute start

3.Check the state of the instance:

virsh list --all
  1. If error is produced: “error: failed to connect to the hypervisor” check the status of libvirtd service and attempt to start if stopped:
service libvirtd status
service libvirtd start

5If instance is not running, reset its state and then attempt to start or stop it:

nova reset-state --active <instance-id>
nova <start/stop> <instance-id>

Note: The above will need to be performed on a controller node or wherever the Nova CLI is accessible. If issue persists or a service is unable to be started, check the following log files on the host:

/var/log/nova/nova-compute.log
/var/log/libvirt/libvirtd.log