Description of problem: This specific environment has a single L3 router with about 200 FIPs. The keepalived processes associated with the router consumes a large amount of CPU. At times the keepalived instances will transition from active to standby and back every minute or so across the L3 agents. There are about 350 other HA routers. This router will also cause overall system stability on the controller nodes (high load average and pacemaker services failures). Disabling this specific router will resolve the apparent issues. Version-Release number of selected component (if applicable): OSP 13 $ grep neutron installed-rpms openstack-neutron-12.1.1-6.el7ost.noarch Sun Apr 12 08:25:13 2020 openstack-neutron-common-12.1.1-6.el7ost.noarch Sun Apr 12 08:21:38 2020 openstack-neutron-l2gw-agent-12.0.2-0.20190420004620.270972f.el7ost.noarch Sun Apr 12 08:26:48 2020 openstack-neutron-lbaas-12.0.1-0.20190803015156.b86fcef.el7ost.noarch Sun Apr 12 08:26:08 2020 openstack-neutron-lbaas-ui-4.0.1-0.20190723082436.ccf8621.el7ost.noarch Sun Apr 12 08:26:07 2020 openstack-neutron-linuxbridge-12.1.1-6.el7ost.noarch Sun Apr 12 08:26:48 2020 openstack-neutron-metering-agent-12.1.1-6.el7ost.noarch Sun Apr 12 08:26:48 2020 openstack-neutron-ml2-12.1.1-6.el7ost.noarch Sun Apr 12 08:21:39 2020 openstack-neutron-openvswitch-12.1.1-6.el7ost.noarch Sun Apr 12 08:26:46 2020 openstack-neutron-sriov-nic-agent-12.1.1-6.el7ost.noarch Sun Apr 12 08:26:49 2020 puppet-neutron-12.4.1-10.el7ost.noarch Sun Apr 12 08:25:11 2020 python2-neutronclient-6.7.0-1.el7ost.noarch Thu Apr 18 10:02:11 2019 python2-neutron-lib-1.13.0-2.el7ost.noarch Sun Apr 12 08:21:24 2020 python-neutron-12.1.1-6.el7ost.noarch Sun Apr 12 08:21:29 2020 python-neutron-lbaas-12.0.1-0.20190803015156.b86fcef.el7ost.noarch Sun Apr 12 08:21:29 2020 How reproducible: Unknown, This specific environment Additional info: I'll provide additional details in private comments
*** This bug has been marked as a duplicate of bug 1869355 ***