The alerting is based on throughput utilization & data peaks in/out. There are future plans to introduce baselining and trends which identify normal patterns and alert on deviations from these baselines.