In case ctl01 node is down, the current mechanism will return
invalid data (the string "Minion") instead of a valid controller IP.
Since "keystonercv3" is present on all controller nodes at </root>,
adjust the credentials fetch to use the first online controller node.
Change-Id: I4ba15bb40d465b112c5e9fc827e3e2a69e671b26
Signed-off-by: Alexandru Avadanii <Alexandru.Avadanii@enea.com>
# retrieving controller vip
controller_ip=$(ssh 2>/dev/null ${ssh_options} ubuntu@${installer_ip} \
- "sudo salt --out txt 'ctl01*' pillar.get _param:openstack_control_address | awk '{print \$2}'" | \
+ "sudo salt --out txt 'ctl*' pillar.get _param:openstack_control_address | awk '{print \$2; exit}'" | \
sed 's/ //g') &> /dev/null
info "Fetching rc file from controller $controller_ip..."