Alertmanager and Grafana
Alertmanager (standalone, or the one bundled with Grafana’s alerting) can call a webhook receiver directly. InfraInbox accepts its native payload, so nothing changes on the Prometheus or Grafana side beyond adding a receiver or contact point.
Tested with: Alertmanager 0.28, Grafana 11+.
What you get
Section titled “What you get”A firing alert opens an incident; the matching resolved notification closes it, matched by Alertmanager’s own alert fingerprint. The shipped preset sets firing alerts to never auto-expire — only the resolved notification is trusted to close them, since Alertmanager already tracks that state for you.
Prerequisites
Section titled “Prerequisites”- Alertmanager (or Grafana’s alerting) able to reach your InfraInbox server over HTTPS.
send_resolved: trueon the receiver, or you’ll never see an incident close on its own.
Alertmanager
Section titled “Alertmanager”-
Go to Add source → Alertmanager and name it. Copy the issued ingest key.
-
Add a receiver to
alertmanager.ymland route alerts to it:receivers:- name: infrainboxwebhook_configs:- url: https://infrainbox.example.com/hooks/alertmanagersend_resolved: truemax_alerts: 500http_config:authorization:credentials: iik_src_… -
Reload Alertmanager and click through a route to
infrainbox, or trigger a real alert to confirm.
Grafana
Section titled “Grafana”-
Alerting → Contact points → Add contact point, integration Webhook.
-
Fill in:
Field Value URL https://infrainbox.example.com/hooks/alertmanagerHTTP Method POST Authorization Header Scheme Bearer Authorization Header Credentials iik_src_… -
Click Test. The source shows Reachable.
-
Point a notification policy at the contact point.
Check it works
Section titled “Check it works”Trigger an alert (or use Grafana’s Test button) and confirm the source switches to Receiving events and an incident opens. Let the alert clear and confirm the incident resolves.
Limitations
Section titled “Limitations”One InfraInbox source can take every Alertmanager or Grafana instance’s alerts you want grouped the same way — split them into separate sources only if you want separate routing or separate presets.
- Routing rules — route by alert label or severity
- Presets — edit the shipped rule