Skip to main content

Grafana Alerts

We currently use Grafana's zero-config out-of-the-box Kubernetes monitoring solution.

See Grafana (Kubernetes).

This comes with several sane default alerts configured by Grafana, eg. for CPU/Memory usage and crash loops.

Because these alerts are auto-configured and can be noisy, we generally assign them low priority (doesn't page), but they shouldn't be ignored, and alerts can be assigned to be high-priority PagerDuty alerts in the Grafana UI.

Viewing/Editing Alerts​

Default Grafana Alerts for Kubernetes

Alerts can be viewed by going to the Grafana Dashboard and clicking on Alerting > Alert rules.

Alerts can also be edited on this page. By default, these alert rules are created and managed by Grafana.

Alert Notifications​

Alert notifications can be edited by going to the Contact points and Notification policies pages.

warning

Make sure the non-default Alertmanager created by the Grafana Kubernetes integration is selected as shown below.

Notification Policies