What teams usually need from an OnCall alternative
Teams evaluating an alternative generally need four things to work together: monitoring integrations, an accurate on-call schedule, a routing model and a reliable way to acknowledge or escalate alerts.
IncidentRelay provides these capabilities as a standalone application. It can receive alerts from Prometheus Alertmanager, Zabbix, Sentry and generic webhooks, then connect them to services, teams, rotations and notification channels.
Alert routing without a required dashboard stack
IncidentRelay does not require Grafana to operate. Alertmanager users can send alerts directly to route-specific endpoints, while other monitoring tools can use dedicated integrations or the generic webhook API.
Routes can match alert labels, select services and teams, apply grouping behavior and attach notification channels. Each route has its own intake token, keeping alert submission separate from outgoing delivery credentials.
Scheduling and escalation
Rotations support layers, membership periods and temporary overrides. Teams can view the final schedule inside IncidentRelay or subscribe through CalDAV and ICS feeds.
Escalation policies define ordered response steps. An unacknowledged incident can trigger reminders and move through additional users or rotations. Responders can also request help from a user, team, rotation or policy during an active incident.
Notification channels
- Mattermost webhook or Bot API notifications.
- Telegram notifications and action handling.
- Email delivery and configurable templates.
- Slack, Discord, Microsoft Teams and generic outbound webhooks.
- Profile-level browser and PWA push.
- Pluggable voice-call providers for custom telephony.
Operational context beyond a notification
IncidentRelay groups related alerts into operational incidents. Responders can acknowledge, resolve, silence and comment on them, set priorities and request additional responders. Services add ownership, runbooks, dependencies, impact information and default stakeholders.
Deployment and ownership
The platform can run with Docker Compose, RPM and systemd deployments or Kubernetes using its Helm chart. SQLite supports smaller single-node installations, while PostgreSQL is recommended for larger or long-running environments.
This deployment flexibility is useful for teams that want the on-call system in the same operational environment as the rest of their internal tooling. It also means the team owns upgrades, backups and availability.
Grafana OnCall migration
Migrate users, teams, memberships, schedules, escalation policies, integrations and alert routes.
The migration tools use public APIs, never connect to product databases and run in dry-run mode by default. Review the generated report before applying changes.
How to evaluate IncidentRelay
Test it with the monitoring source and notification channel you actually use. Build one rotation, send one firing alert, acknowledge it, send a resolved payload and verify the escalation behavior. A realistic test exposes workflow differences much faster than comparing screenshots.
IncidentRelay is actively developed. Review the release history before planning a production rollout.
Grafana and Grafana OnCall are trademarks of Grafana Labs. IncidentRelay is not affiliated with Grafana Labs.