Description of problem:

Having several nodes on OSP8 and trying to introspect them.
This fails if there is no DISPLAY on the server hosting the VMs.

Nodes were defined this way:

                . ./stackrc ; \
                ironic node-create -n ${IRONIC_NODE} \
                -d pxe_ssh \
                -i ssh_address=${VBOX_HOST_IP} \
                -i ssh_username=${VBOX_USER} \
                -i ssh_virt_type=vbox \
                -i ssh_password=\"${VBOX_USER_PWD}\" \

                . ./stackrc ; \
                ironic node-update ${IRONIC_UUID} add \
                properties/cpus=${vm_slave_cpu_default} \
                properties/memory_mb=${vm_slave_memory_default} \
                properties/local_gb=62 \
                properties/cpu_arch=x86_64 \
                \""

'openstack introspection bulk start' hung forever and In the logs I was able to notice these outputs:
 Error: Failed to execute command via SSH: LC_ALL=C /usr/bin/VBoxManage startvm osp-baremetal-9.", "updated_at": "2016-03-24T15:08:40+00:00", "maintenance_reason": null, "provision_state": "manageable", "clean_step": {}, "uuid": "71392099-f84c-449e-88a1-e3b1fd1af248", "console_enabled": false, "target_provision_state": null, "provision_updated_at": "2016-03-24T14:24:18+00:00", "power_state": "power on", "inspection_started_at": null, "inspection_finished_at": null, "maintenance": false, "driver": "pxe_ssh", "reservation": null, "properties": {"memory_mb": 4096, "cpu_arch": "x86_64", "local_gb": 62, "cpus": 2}, "instance_uuid": null, "name": "osp-baremetal-9", "driver_info": {"ssh_username": "raistlin", "ssh_virt_type": "vbox", "ssh_address": "10.0.128.244", "ssh_password": "******"}, "created_at": "2016-03-24T14:23:08+00:00", "ports": [{"href": "http://10.20.0.2:6385/v1/nodes/71392099-f84c-449e-88a1-e3b1fd1af248/ports", "rel": "self"}, {"href": "http://10.20.0.2:6385/nodes/71392099-f84c-449e-88a1-e3b1fd1af248/ports", "rel": "bookmark"}], "driver_internal_info": {}, "instance_info": {}}, {"target_power_state": null, "links": [{"href": "http://10.20.0.2:6385/v1/nodes/031d9a8e-f8fe-41bb-a41c-17be0519088d", "rel": "self"}, {"href"

Sure enough, I was able to reproduce this at the CLI:
$ VBoxManage startvm osp-baremetal-8
Waiting for VM "osp-baremetal-8" to power on...
VBoxManage: error: The virtual machine 'osp-baremetal-8' has terminated unexpectedly during startup with exit code 1 (0x1)
VBoxManage: error: Details: code NS_ERROR_FAILURE (0x80004005), component MachineWrap, interface IMachine

The workaround is to invoked startvm with '--type headless':
$ VBoxManage startvm --type headless osp-baremetal-8
Waiting for VM "osp-baremetal-8" to power on...
VM "osp-baremetal-8" has been successfully started.
Version-Release number of selected component (if applicable):


How reproducible:

100% if none is logged at the console (no DISPLAY running)


Additional info:
stack@instack ~]$ ironic node-list
+--------------------------------------+------------------+---------------+-------------+--------------------+-------------+
| UUID | Name | Instance UUID | Power State | Provisioning State | Maintenance |
+--------------------------------------+------------------+---------------+-------------+--------------------+-------------+
| 4caf2c9d-1d04-4863-b4ae-810e560bcbcc | osp-baremetal-1 | None | power off | available | False |
| 5f0c0f63-53fa-4f13-9d4c-4565eb8b75e4 | osp-baremetal-2 | None | power off | available | False |
| 53bc95dd-bf1a-49ec-ae22-518edabfe019 | osp-baremetal-3 | None | power off | available | False |
| 61065f76-0160-4ca4-873b-e3b6da944e5e | osp-baremetal-4 | None | power off | available | False |
| 82b73be8-d1b2-49d0-811d-751c58722a0f | osp-baremetal-5 | None | power off | available | False |
| c8cd6846-1b09-427e-95b4-0cb7a4aafc3e | osp-baremetal-6 | None | power off | available | False |
| 30c63703-13d7-4b82-8c22-c83e972fb7cd | osp-baremetal-7 | None | power off | available | False |
| 74a19eb4-1c90-4afb-8b8b-b5f51be9d9ef | osp-baremetal-8 | None | power off | available | False |
| 71392099-f84c-449e-88a1-e3b1fd1af248 | osp-baremetal-9 | None | power off | available | False |
| 031d9a8e-f8fe-41bb-a41c-17be0519088d | osp-baremetal-10 | None | power off | available | False |
| 927b3514-6d6f-4d80-af00-47442c266871 | osp-baremetal-11 | None | power off | available | False |
| 18e66816-c35b-4473-a4e9-df8d3c6f4fad | osp-baremetal-12 | None | power off | available | False |
+--------------------------------------+------------------+---------------+-------------+--------------------+-------------+
[stack@instack ~]$ ironic node-show 74a19eb4-1c90-4afb-8b8b-b5f51be9d9ef
+------------------------+----------------------------------------------------------------------+
| Property | Value |
+------------------------+----------------------------------------------------------------------+
| target_power_state | None |
| extra | {u'hardware_swift_object': u'extra_hardware-74a19eb4-1c90-4afb-8b8b- |
| | b5f51be9d9ef'} |
| last_error | None |
| updated_at | 2016-03-24T16:35:37+00:00 |
| maintenance_reason | None |
| provision_state | available |
| clean_step | {} |
| uuid | 74a19eb4-1c90-4afb-8b8b-b5f51be9d9ef |
| console_enabled | False |
| target_provision_state | None |
| provision_updated_at | 2016-03-24T15:11:34+00:00 |
| maintenance | False |
| inspection_started_at | None |
| inspection_finished_at | None |
| power_state | power off |
| driver | pxe_ssh |
| reservation | None |
| properties | {u'memory_mb': u'4096', u'cpu_arch': u'x86_64', u'local_gb': u'62', |
| | u'cpus': u'2'} |
| instance_uuid | None |
| name | osp-baremetal-8 |
| driver_info | {u'ssh_username': u'raistlin', u'ssh_virt_type': u'vbox', |
| | u'ssh_address': u'10.0.128.244', u'ssh_password': u'******'} |
| created_at | 2016-03-24T14:22:57+00:00 |
| driver_internal_info | {} |
| chassis_uuid | |
| instance_info | {} |
+------------------------+----------------------------------------------------------------------+