Anomaly detection alert
Home > Select Project > Alert > Event Setting > Anomaly Detection tab
Anomaly detection is a set of behaviors that look for unexpected patterns in data such as anomaly, outlier, and abnormal. Based on the metrics data, you can set alerts to be sent when the rising and falling patterns are not as usual.
By selecting + Add alert policy, you can perform the settings for the anomaly detection event. After all settings are finished, select Save. By selecting the toggle button on the right of Anomaly Detection, you can select the activation.
Creating the anomaly detection event
Event basic information
Enter the basic data for the anomaly detection event.
-
Level: Select a level among Info, Warning, and Critical.
-
Message: Enter an alert message to be displayed when an event occurs. You can apply variables to the message by entering
${tag or field key}
. The key to enter into the variable must be a value contained in the selected metric data, Category.
Metrics
Select an event occurrence target.
-
Category: The unit to identify the metrics data. It is a mandatory value for setting the metrics events.
-
Field: Select a field to use as an event generation condition. You can select multiple items.
-
Filter: Select an event condition target. You can select multiple items.
-
Object Merge: You can select an object merging method.
Anomaly Detection
You can select the severity of Rising Pattern or Falling Pattern. This function can be enabled or disabled by selecting the toggle buttons for each item.
Modifying the anomaly detection event
-
Go to Alert > Event Setting and then select the Anomaly Detection tab.
-
In the anomaly detection list, select on the right of the item to modify.
-
The anomaly detection setting window appears. After modifying the desired items and then select Save.
Deleting the anomaly detection event
In Alert > Event Setting, select the Anomaly Detection tab. Select on the right of the item to delete. When the confirmation message appears, select Delete.