Description of problem: - Customer reported an issue where floating IPs assigned to instances were no longer working, but other IPs from the same network range assigned directly were working. Environment: - ML2/OVS - Non-DVR w/ HA router Version-Release number of selected component (if applicable): RHOSP 16.1.6 openstack-neutron-15.2.1-1.20210409073445.40d217c.el8ost.noarch [1] openstack-neutron-server (https://catalog.redhat.com/software/containers/rhosp-rhel8/openstack-neutron-server/5de6be9fd70cc51644a57114?tag=16.1.6-8.1627296179&push_date=1627500607000&container-tabs=packages) How reproducible: N/A Steps to resolve: - Disabled the router - Manually set the router to non-ha in the ovs_neutron database - Enabled the router so it would come up on one controller (from here we could ping the GW and attached FIP) - Disabled the router - Enabled HA mode - Enabled the router (at this stage the router came up on all 3 controllers with one active) Additional info: - Checking qrouter namespaces on controllers via `ip netns list` shows qrouter namespaces are missing. - Can create new routers and the qrouter namespaces correctly get created. - Query `select * from ha_router_agent_port_bindings where router_id="${routerID}"` in `ovs_neutron` DB returned 0 results. - See private comment for logs.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Moderate: Red Hat OpenStack Platform 16.1.9 (openstack-neutron) security update), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2022:8870