Forecasts enable dynamic thresholds removing the need for manual configuration.

Forecasting

InfraSonar stores time series data in SiriDB a database specifcly created for effeciant storing of large volumes of time series data.

As this time series data is not condensed over time it is a great source for forecasting algorithms. InfraSonar has a dedicated mechanism for creating these forecast using historical time series data.

Dynamic thresholds

Using forecasts allows InfraSonar to offer dynamic thresholds as the forecast provides an upper and lower expected value.

These expected values can be used in conditions allowing to alert when a measured value is out of the expected bounds.

Example

One easy example where dynamic thresholds come in handy are volulumes filled with static archive data that exceed the recommended 80% in use threshold. In this scenario a dynamic threshold can be used to detect any changes joined with a static fixed threshold acts as a safeguard avoiding 100% volume full.

This is useful as now one single condition can be used for all volumes avoiding the need for manually managing separated conditions per volume.