It would be good to have events generated for connection attempt failures, similar to the connection events for successes. In some cases, runaway processes may drive resource utilization, and the best response will be to use IP filtering or similar tools to block the traffic. Events capturing that data would provide the infomration necessary to make decisions.