Self-Hosted Observability Directory

Grafana OnCall

Open-source on-call and incident-response — schedules, escalation policies, ChatOps. The closest OSS analog to PagerDuty.

LicenseAGPL-3.0
Repositoryhttps://github.com/grafana/oncall
GitHub stars (approx)3500
LanguagePython / TypeScript
Self-host deploymentdocker-compose, kubernetes helm
Commercial hosted optionyes (vendor-run)
Freshness statusFresh
Freshness checked2026-05-17
Categoriesalerts
Replaces SaaSpagerduty

Grafana OnCall is an open-source incident-response platform: on-call schedules with rotations and overrides, escalation policies, integrations with Grafana, Prometheus Alertmanager, Zabbix, Datadog (as a source), Slack/MS Teams/Telegram, and a mobile app for push alerts. It is the closest 1:1 OSS analog to PagerDuty’s core on-call workflow.

Pick it if: you want PagerDuty’s on-call ergonomics — rotating schedules, multi-step escalation, mobile push — without per-responder seat pricing. Works well bolted onto a Prometheus + Alertmanager stack: Alertmanager routes the alert, OnCall manages who gets woken at 2 a.m. and how it escalates.

Skip it if: you only need a notification router (Alertmanager alone is enough) or you have no on-call rotation to manage. The product surface area is intentionally larger than Alertmanager and brings more operational overhead.

Self-host notes: docker-compose for a quick stand-up; the Helm chart is the supported production path. Backed by Postgres + Redis + Celery workers; budget a small Kubernetes namespace or a 2 GB VM.

Trademark note: “PagerDuty” is a trademark of PagerDuty, Inc. Grafana OnCall is an independent open-source project from Grafana Labs and is not affiliated with or endorsed by PagerDuty, Inc.

← Back to directory