After replacing all ceph-mon units on my bionic/queens openstack cloud I faced the same issue on a instance.
openstack shelve $instance_uuid openstack unshelve $instance_uuid
was a workaround that brought the instance back to life
After replacing all ceph-mon units on my bionic/queens openstack cloud I faced the same issue on a instance.
openstack shelve $instance_uuid
openstack unshelve $instance_uuid
was a workaround that brought the instance back to life