ALERTING.md

Alerting

In order to make potential problems inside the platform visible, alerts can be configured, based on the available metrics.

Approach

Since applications should know best about there failure indicators, the alerts will be configured as a part of the appliction. For that, create a PrometheusRule resource:

    apiVersion: monitoring.coreos.com/v1
    kind: PrometheusRule
    metadata:
        name: prom-rule
    spec:
        groups:
        - name: group.alerts
            rules:
            - alert: AlertName
              expr: |
                <alert experession>
              labels:
                severity: <alert-severtiy>

Please check the general documentation of alerts and the initial keycloak-alert.