Octavia o-hm0 interface can't wait DHCP

[Victor Chembaev]

On huge deployments (2k+ network ports per network node) neutron-openvswitch-agent takes tens of minutes to finish the sync after the host has been rebooted, so, octavia-interface.service systemd unit can't wait to be started (it can't obtain ip address from DHCP, because there are no octavia-lb-network synchronized still on the host) in default 90s (default for DefaultTimeoutStartSec) and fails to start.

Our experience - on deployments up to 3k ports per network node full sync takes up to 30mins. Do more 3k ports is a bad practise (there are redhat recommendations). So we are propose to add

[Service]
TimeoutStartSec=1800

options into the octavia-interface.service.j2 file for Octavia, it will solve problem described

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.opendev.org/c/openstack/kolla-ansible/+/920377