Created attachment 860222 [details] the nova logs Description of problem: while trying to run instance migration on an instance, the migration failed because of incorrect ssh key authentication between the nova compute nodes. After the failed migration: 1. The instance failed to run. 2. Couldn't be deleted. 3. Its volumes couldn't be detached, snapshoted or copied from. Thus all of the consistent data was not available. Version-Release number of selected component (if applicable): python-nova-2014.1-0.11.b2.fc21.noarch openstack-nova-api-2014.1-0.11.b2.fc21.noarch openstack-nova-cert-2014.1-0.11.b2.fc21.noarch openstack-nova-conductor-2014.1-0.11.b2.fc21.noarch openstack-nova-scheduler-2014.1-0.11.b2.fc21.noarch openstack-nova-compute-2014.1-0.11.b2.fc21.noarch python-novaclient-2.15.0-1.fc20.noarch openstack-nova-common-2014.1-0.11.b2.fc21.noarch openstack-nova-console-2014.1-0.11.b2.fc21.noarch openstack-nova-novncproxy-2014.1-0.11.b2.fc21.noarch How reproducible: 100% Steps to Reproduce: 1. launch an instance 2. try to migrate it. 3. see instance status. Actual results: the instance is not responding to any action. Expected results: if the migration fails before doing anything to the instance's files the instance should be up and running & the an error about the failed migration should appear. Additional info:
With current (RDO Juno) packages I am not able to reproduce this behavior. An attempt to live-migrate an instance will "successfully fail" in the face of: - Bad host keys - Bad ssh keys - Bad hostnames - Bad account configuration Whereby "successfully fail" I mean that the instance remains in ACTIVE state and functioning correctly on the original host.