Ok, we were trying to test out Kolide, and the basic complaint I got from our ops is that enrolled OSQuery agents were not shutting up about losing connection when Kolide was down. I basically just want to limit how much it logs about the loss of connection and ensure scheduled queries are still running fine.