Always Firing Alerts

This page documents alerts that are expected to always be firing under normal circumstances. These alerts serve as a heartbeat to verify that the alerting pipeline is functioning correctly.

Purpose

The Watchdog alert is intentionally designed to fire continuously. If it stops firing, this indicates a potential issue with the Prometheus alerting system itself — such as broken rule evaluation or a disconnected alertmanager.

To prevent unnecessary noise, this alert is configured to route to a null receiver in Alertmanager, meaning it will not generate notifications.

Alert Definition

Alert	Severity	Trigger After	Description	Threshold Ref
Watchdog	NONE	0m	This alert should always be firing. If it stops, the alert pipeline may be broken.