Our undercloud initial install was failing with queens because the
--use-heat arg does not work the same way as master/rocky. Therefore the
undercloud install would fail, and then try to reinstall on the next
ansible task. This one would succeed, via container installation because
we were using the wrong undercloud/overcloud disk images.
Change-Id: Ifb829aeeec991ba6dbe582fda3e9ee6b9bc733df
Signed-off-by: Tim Rozet <trozet@redhat.com>
# Patch applies to overcloud/undercloud
if docker_image is None:
oc_url = urllib.parse.urljoin(
- con.UPSTREAM_RDO.replace('master', branch), 'overcloud-full.tar')
+ con.UPSTREAM_RDO.replace(con.DEFAULT_OS_VERSION,
+ branch), 'overcloud-full.tar')
oc_mtime = utils.get_url_modified_date(oc_url)
if oc_mtime > submitted_date:
logging.debug("oc image was last modified at {}, which is"
DEBUG_OVERCLOUD_PW = 'opnfvapex'
NET_ENV_FILE = 'network-environment.yaml'
DEPLOY_TIMEOUT = 120
-UPSTREAM_RDO = 'https://images.rdoproject.org/master/delorean/current' \
- '-tripleo-rdo/'
+UPSTREAM_RDO = 'https://images.rdoproject.org/{}/delorean/current' \
+ '-tripleo-rdo/'.format(DEFAULT_OS_VERSION)
OPENSTACK_GERRIT = 'https://review.openstack.org'
DOCKER_TAG = 'current-tripleo-rdo'
when: aarch64
- block:
- name: undercloud install
- shell: openstack undercloud install --use-heat False &> apex-undercloud-install.log
+ shell: openstack undercloud install &> apex-undercloud-install.log
become: yes
become_user: stack
rescue: