Provides a computation and storage framework for aggregated metrics. In most systems, metrics will be initially recorded as events. Using the underlying time-series database Clickhouse this can be efficiently queried and used for intense computations.
It is still advisable to compute and store some aggregations (number of X per month per entity) so that they can be readily displayed in dashboards without much computation.
This framework therefore provides a database independent storage called Metrics which has an implementation for JDBC and MongoDB.
One of these frameworks (biz.analytics-metrics-jdbc or biz.analytics-metrics-mongo) has to be enabled.
To compute (aggregate) metrics for entities, DailyMetricComputer or MonthlyMetricComputer can be subclassed. Note that monthly metric computers are scheduled in a best effort manner on a daily basis for the current month so that these aggregations are also available if possible.
Additionally, global metrics can be computed by subclassing DailyGlobalMetricComputer or MonthlyGlobalMetricComputer.
Note that these schedulers are managed by the scheduling system found in the scheduler package.