This article provides an overview of events, and includes the following:
What are Events?
Events are points in time worth mentioning. In contrast to metric time series data, events can happen at irregular intervals, can be based on calendars, or even be ad-hoc occurrences.
Some event examples:
- Holidays by country
- Lunar calendar events (such as Muslim holidays)
- Lunisolar calendar events (Chinese, Japanese, Jewish holidays, and more)
- Sales and marketing events (Black Friday, Autumn Sale, and more)
- Weather events (such as blizzard days, heavy rains, hurricanes, and more)
- Sporting events
- Financial events (such as the publication of Interest rates, payment days, bank closure days, and end-of-month accounting days)
- Technical events (such as technological deployment events)
How to use Events
Use events to create a timeline, such as deployment events, holidays and alerts that occur periodically, yet that may impact the behavior of monitored metrics.
An event or set of aggregated events is plotted as a single vertical line on a chart. Each event has a time and a meta-data description. Historical and future events can be programmed; to program events, see the Anodot APIs Reference - Events.
There are four main types of events in Anodot:
- Display events: The default displayed events in Anodot.
- Influencing events: These events adjust the metric baseline near a specific event, such as Christmas sales or traffic periods. Over time, with historical data assisting event accuracy, they help reduce anomalies. For more information, see Influencing Events.
- Suppress events: These events are used to suppress specific metrics from an alert. For more information, see Suppressing Events.
- Office hours events: These events enable you to pause alert triggers during a specified period. For more information, see Office Hours Events.