Introduction

OnCallAlerting is on-call scheduling and alert escalation for teams that already have monitoring. It does one job: take an alert from the monitoring you already run, work out who is on call, tell them, and keep escalating until someone acknowledges.

It is not a monitoring tool. Your alert rules stay where they are. You point their webhook, or their PagerDuty or Opsgenie integration, at OnCallAlerting.

How an alert moves through OnCallAlerting

1. An alert source receives an alert. Each source is a webhook URL or an integration key, tied to one escalation policy.
2. The alert opens an incident, or adds an occurrence to the incident that already has the same deduplication key.
3. The escalation policy runs. Level 1 is notified straight away. Each later level is notified if nobody acknowledges during the wait before it.
4. Rosters decide who that is. A level can target rosters and people. A roster notifies whoever is on call at that moment, including substitutions and follow-the-sun shifts.
5. People are notified in their in-app inbox, and in Slack when the level uses Slack.
6. Someone acknowledges. Escalation stops at any level, in any round. When the monitoring tool sends a recovery, the incident resolves.

The parts

Part	What it is
Organization	Everything belongs to one organization: people, rosters, policies, sources, heartbeats and incidents. The person who signs up is its primary owner.
Roster	Who is on call and when. A simple rotation, several layers, or follow-the-sun working shifts.
Escalation policy	Ordered levels. Each level notifies rosters and people, then waits.
Alert source	Where alerts come in. Native payloads, a generic JSON webhook, or PagerDuty- and Opsgenie-compatible APIs.
Incident	One problem, with its status, occurrences, notes and activity timeline.
Heartbeat	A scheduled job that should ping OnCallAlerting. A missed ping opens an incident.
Outbound webhook	Signed incident events sent to your own tools.

Roles

Role	Can do
Primary owner	Everything an administrator can. Receives the page when an escalation level finds nobody on call.
Administrator	Manage rosters, substitutions, escalation policies, alert sources, heartbeats, Slack, outbound webhooks and invitations.
Member	See the workspace, acknowledge and resolve incidents, and take every incident action: notes, priority, reassign, snooze, merge and responders.

See members and roles for the full list.

What OnCallAlerting does not do yet

Check this before you rely on it at night

Notifications are in-app and Slack only. OnCallAlerting does not send email, SMS, voice calls or mobile push, and there is no mobile app. Neither channel can wake someone who is asleep, so it fits on-call that happens during working hours or follows the sun.

• No import of schedules or policies from another on-call tool.
• No SSO, SAML or SCIM.
• No API keys for your own scripts. Alerts come in through alert sources and heartbeats; events go out through outbound webhooks.
• No way yet to change a teammate's role or remove them from the organization in the app.
• No status pages.

Next

Follow the quick start to go from a new organization to an acknowledged test alert.