Configuring Alerts
When an anomaly is detected, the system generates a notification called an alert. Based on the statistics pegged by the system, alerts are fired. You can configure an expression to fire an alert when the expression becomes true.
The CEE uses the Prometheus Alert Manager for alerting operations. The CEE YANG model - either through CLI or API - allows users to view the active alerts, silenced alerts and alert history. A predefined set of alerting rules are added whenever the application is installed or updated. Also, the applications can call the alert API directly to add or clear alerts. The Prometheus Alert Manager API (v2) is the standard API used.
The Prometheus Alerts Manager includes the following options:
-
Defining Alert Rules – This option defines the Alert Manager on what to alert. You can define the alerts using the Prometheus Query Language (PromQL).
-
Defining Alert Routing – This option defines the Prometheus Alert Manager on what to do with the received alerts. At present, the SNMP Trapper is supported as the outbound alerting. Also, the CEE provides an Alert Logger for storing the generated alerts.