Move beyond a shared alert inbox
An alert can reach a team inbox while everyone assumes someone else is handling it. On-call schedules make responsibility explicit, and escalation defines what happens when the first page has not been acknowledged.
PageLantern brings these steps into the monitoring workspace. Choose who covers each rotation, where personal notifications should go and which incident types should start escalation. Existing workspace notification rules remain separate, so review both paths when deciding who should receive an alert.
Build a rotation around your team’s handoff time
An owner or admin can open Alerting → On-call to create a schedule with a time zone, local start time, rotation length and ordered responders. Add a backup and use temporary overrides for holidays or other short-term changes.
Calendar-day rotations preserve the local handoff time across daylight-saving changes. The schedule exposes coverage gaps, including unavailable responders. A filled calendar shows who is assigned; it does not prove that their contact details work or that a notification was delivered.
Verify contacts before the first incident
Each responder adds their own email address or international SMS number under Settings → Notifications → My on-call contacts. Verify the destination and test it before relying on it. Changing a destination removes its verification.
SMS verification, tests and incident messages use workspace SMS credits. Delivery also requires configured email or SMS transport. This workflow supports email and SMS; plan contact coverage around those channels and check the readiness information shown in the workspace.
Define the next step before it is needed
Under Alerting → Escalation, select the monitors, incident severities and schedule, then add up to eight stages. A stage can contact the current primary, the backup or named responders by email, SMS or both. Delays increase from the time the incident episode opened.
Use Preview to inspect current assignments without sending messages. Start with a small, understandable policy and make sure each stage has a reachable responder. The incident view shows assignments, the next escalation and delivery history when the policy runs.
Distinguish taking responsibility from restoring service
Acknowledge and stop escalation ends paging while keeping the incident open. Hand off to another responder assigns a verified responder, clears the earlier acknowledgement and queues a new page. Neither action claims the monitored service has recovered.
Provider acceptance means a message was accepted for delivery, not that someone read it. Exhausted stages mean there is no further stage queued. Use those distinctions during an incident, and review the delivery history afterward to improve the policy.