Skip to content

Update PrometheusNotConnectedToAlertmanagers alert to trigger after 30m

Pierre Guinoiseau requested to merge pguinoiseau/extend-prometheus-not-connect-alert-threshold into master

The Prometheus pods can take over 15 minutes to restart now, so this alert would fire even during upgrades or node rotations. Extending to 30 minutes is not ideal but it will reduce the number of alerts to actual incidents when pods are unable to start.

Merge request reports