To Reproduce Steps to reproduce the behavior:

1. Redeploying an existing hypervisor with the same hostname results in nova-compute startup error without truly confirm.

Expected behavior

•  The documentions can teach user how to verified the hypervisor status before redeploying with same hostname

Bug impact

• It seems the commiunity already confirm this bug
• https://bugs.launchpad.net/kolla-ansible/+bug/2051011

Known workaround

• Get the ID of the hypervisor which down`
  openstack hypervisor list |grep down
• Add this ID to /var/lib/nova/compute_id (in the case you use local storage)

• sudo -i
  echo "<The UUID from first step>" > /var/lib/nova/compute_id
  chown 42436:42436 /var/lib/nova/compute_id
  chmod 644 /var/lib/nova/compute_id  

Additional context