The route configuration controls how the alerts are grouped and filtered before they reach the receiver. Prometheus uses routes to manage and control the grouping, filtering, and delivery of alerts.

• Grouping: Group alarms of the same type to avoid sending a large number of similar alerts repeatedly. You can set the sending interval to ensure that alarm messages are not too frequent, and aggregate similar alarms to reduce repeated notifications. For example, if the CPU usage of multiple nodes exceeds the threshold, these alarms can be aggregated into one notification.
• Filtering: You can filter alert messages based on specific labels or rules, and only send an alert that meet the conditions. For example, only alerts related to a service are sent to a specific administrator group. This helps reduce irrelevant or low-priority alerts.
• Retry interval: Set the retry interval and notification frequency when the alarm fails to be sent to avoid excessively resending of alarms.

Admin users can send alerts to recipients through Default Route. Default Route can handle unclassified alarms, but customized Routes provide a more flexible and controllable alarm management method. By setting custom routes, you can refine the processing, filtering, grouping, and sending of alarms according to different needs, thereby ensuring that the right people receive appropriate notifications at the right time, improving the efficiency and accuracy of alarm processing.

Note:

Configure a receiver before creating a route.

Procedure

1. Go to EonKube > Monitoring > Alerting > Routes and Receivers.
2. Click the Routes tab.
3. Click Create.
4. For the Receiver, select a receiver.
5. Go to Grouping and specify the group information:
   • Group by: Specify a tag. Specify the label by which incoming alerts are grouped together. The tag must be unique. If you specify the special tag "..." (aggregate all possible tags), the tag must be the only element in the list. A list of strings is allowed. The default value is N/A.
   • Group Wait: Specify how long for Kubernetes to wait to buffer alerts of the same group before initially sending an alert. The default value is 30s.
   • Group Interval: Specify how long for Kubernetes to wait before sending an alert that has been added to a group of alerts for which an initial notification has already been sent. The default value is 5m.
   • Repeat interval: Specify how long for Kubernetes to wait before sending a repeated alert again. The default value is 4h.
6. Go to Matching. Click Add match and enter a key-value pair to specify the Match information. This setting indicates a set of equality matchers. The matchers are used to identify which alerts to send to a given Route based on labels defined on that alert.
7. Click Create.