Description of problem:
When OpenShift node runs on OpenStack, it can be notified about it by setting an authentication URL and creds. In that case it is trying to guess the instance id of the VM by searching for instance with instance name same as the hostname. There however can be more of such instances if more of same (heat) stacks of openshift are running under same tenant (while running in completely different networks). For example a staging and production environment.

There is a better way to get the instance id -- use the uuid from the metadata :

[cloud-user@ose3-master-h57q6 ~]$ curl 'http://169.254.169.254/openstack/latest/meta_data.json' | json_reformat 
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  1711  100  1711    0     0  21236      0 --:--:-- --:--:-- --:--:-- 21387
{
    "random_seed": "...",
    "uuid": "44447f54-fed0-4d4a-9fed-b7cae0b610ce",
    "availability_zone": "nova",
    "keys": [
        {
            "data": "...",
            "type": "ssh",
            "name": "jenkins"
        }
    ],
    "hostname": "ose3-master-h57q6",
    "launch_index": 0,
    "public_keys": {
        "jenkins": "ssh-rsa ..."
    },
    "name": "ose3-master-h57q6"
}


Version-Release number of selected component (if applicable):
kubernetes.x86_64             1.2.0-0.13.gitec7364b.el7    rhel-extras-candidate
kubernetes-client.x86_64      1.2.0-0.13.gitec7364b.el7    rhel-extras-candidate
kubernetes-master.x86_64      1.2.0-0.13.gitec7364b.el7    rhel-extras-candidate
kubernetes-node.x86_64        1.2.0-0.13.gitec7364b.el7    rhel-extras-candidate


How reproducible:
always


Steps to Reproduce:
1. Boot the stack, rename the instance to some name.
2. start the atomic-openshift-node
3.


Actual results:
# journalctl -xe
zář 13 01:23:19 ose3-master-h57q6 atomic-openshift-node[50596]: I0913 01:23:19.645808   50596 openstack.go:289] Claiming to support Instances
zář 13 01:23:19 ose3-master-h57q6 atomic-openshift-node[50596]: E0913 01:23:19.695103   50596 kubelet.go:1131] Unable to construct api.Node object for kubelet: failed to get external ID from cloud provider: Failed to find object


Expected results:
No complains about the failure to get external ID.


Additional info: