The RDBMS implementation used for storing events and general CM state is not well-suited for fine-grained, timeseries data.

It will be keeping track of high-level stats (total checks runs, total events checked, etc...) for an all-time frame and an operator-configurable frame (recommended daily) -- but that's it. Instead of fleshing out super fine-grained stats for everything we will implement optional output to influxDB.