Description of problem: we are currently getting a constant failover of some of the loadbalancers. this seems related to the number of members in the pool reducing the pool members seems a workaround for this issue Version-Release number of selected component (if applicable): RHOSP 16.1.9 ML2-OVN How reproducible: CU is able to reproduce this, but it is specific to their env Actual results: loadbalancers keep failovering continuously the amphora Expected results: loadbalancer amphora doesn't get failovered a detailed description of the behavior will follow as private comment