Description of problem: When using machines where the DHCP (management) is configured on eth1 and eth0 is meant to be used as the provisioning network for osp-d, even when supplying "eth0" in the "local_interface", my eth1 interface was still replaced with params meant to be used for the provisioning (attached to the OVS bridge and such) Steps to Reproduce: 1. Setup a system where the management interface is not the first in the list 2. Configure the undercloud.conf with "openstack-config --set undercloud.conf DEFAULT local_interface eth0" 3. Run undercloud installation Actual results: Machine looses connectivity as the management interface is being overwritten Expected results: Installation works on the given interface
This bug did not make the OSP 8.0 release. It is being deferred to OSP 10.
Is this problem still reproducible with more recent OSP versions?