Delivery for Prometheus and Grafanas
Alert delivery is handled by Alertmanager, which receives alerts from Prometheus and decides who is notified. It groups related alerts, deduplicates them, applies silences and inhibition, and routes them by label to receivers such as email, Slack or PagerDuty, using a routing tree.
Stage: Alerting · Part of the Prometheus and Grafana roadmap · Last reviewed 19 November 2026
What to learn in Delivery
- Configure a route tree and receivers in alertmanager.yml
- Group alerts by labels to reduce notification storms
- Create a silence for planned maintenance
- Use inhibition so a root-cause alert mutes dependent ones
- Test an end-to-end alert from rule to receiver
Delivery interview questions
What does Alertmanager do that Prometheus does not?
Prometheus evaluates rules and fires alerts. Alertmanager takes those alerts and handles grouping, deduplication, silencing, inhibition and routing to notification channels, and can run as a cluster for high availability.
What is alert inhibition?
Inhibition suppresses certain alerts while another, usually more fundamental, alert is firing. For example, if a whole datacentre is down, per-service alerts from it are muted so responders see the root cause first.
Where Delivery fits in the Prometheus and Grafana roadmap
Delivery belongs to the Alerting stage (intermediate level). Know when to act.
Other units in this stage
Back to the full Prometheus and Grafana roadmap