Alerts

This page explains how to receive alerts from your application when something goes wrong.

How it works

Stakater operates a single Alertmanager that handles alerts from both metrics and logs. You declare rules and routing through your application's values.yaml for the Stakater Application Helm Chart; the platform evaluates the rules and routes the resulting alerts to the channel you configure. You do not run Alertmanager, configure routing trees, or manage receivers.

The platform also ships a curated set of predefined PrometheusRules for common Kubernetes failure modes (pod crash-looping, deployment replica mismatch, PVC filling up, and so on). You can route these to your channels — there is no need to write the rules yourself.

What you do

1. Route alerts from your namespace to a receiver by declaring an AlertmanagerConfig through the Application Chart.
2. Optionally declare custom PrometheusRule or AlertingRule (Loki) resources for conditions the predefined rules do not cover.

Both pieces are scoped to your namespace; Alertmanager merges your routing into its base configuration automatically.

Next step

Continue to Configure application alerting to wire your first alert.