About Forecasting

Site24x7's forecasting engine enables you to predict future points of an AWS service performance metric (measurement of resource usage) based on historical observations.

Dataset: The number of training days depends on each monitor type of AWS. To predict what your metric usage will be in the next seven days, it takes about fifteen to thirty days of historical data.

Accuracy: Site24x7 uses machine learning and traditional time series forecasting models like exponential smoothing to predict future values. To choose the right model, the time series data is broken down into systematic and unsystematic components like level, trend, seasonality and, noise.

Limitations: You can only view forecasts as charts, you can't configure thresholds for the predicted values.

How to view forecast charts?

Sign in to the Site24x7 web console.
Choose AWS from the left navigation pane and choose the monitored AWS account
In the menu drop, choose the AWS service (list of supported services and corresponding performance metrics are listed below), and choose the instance of that service for which you want to view forecast charts.
Choose the Forecast tab to view the charts.

View the AWS performance metrics and enable AWS forecasting

AWS metric forecasting is available for the following services: