The service level availability of some of the entries may degrade over time, they may not see regular updates or they may be plagued by unfixed configuration issues that compromise their ability to fully participate in the federation.
Independent, third party automation could keep checking the health of each according to standardized criteria and alert the list maintainers and the given relay operator in case of issues. The list maintainers should either mark the entries that never have problems as suggested or remove problematic ones. As an alternative, the measured metrics and the list of issues found could be summarized in a table automatically.
Via: