Periodically checks the health of all clusters registered on the router and updates each node status accordingly. When initialized, it schedules a task to call each node's health endpoint and marks nodes as UP or DOWN based on the HTTP response. This ensures the load balancer always has up-to-date status for routing decisions.