Paging and notifications (FREE)
When there is a new alert or incident, it is important for a responder to be notified immediately so they can triage and respond to the problem. Responders can receive notifications using the methods described on this page.
Introduced in GitLab 13.1.
Responders can be paged via Slack using the Slack Notifications Service, which you can configure for new alerts and new incidents. After configuring, responders receive a single page via Slack. To set up Slack notifications on your mobile device, make sure to enable notifications for the Slack app on your phone so you never miss a page.
Email notifications for alerts
Email notifications are available in projects for triggered alerts. Project members with the Owner or Maintainer roles have the option to receive a single email notification for new alerts.
- On the top bar, select Menu > Projects and find your project.
- On the left sidebar, select Settings > Monitor.
- Expand Alerts.
- On the Alert settings tab, select the Send a single email notification to Owners and Maintainers for new alerts checkbox.
- Select Save changes.
Update the alert's status to manage email notifications for an alert.
In projects that have an escalation policy configured, on-call responders can be automatically paged about critical problems through email.
Escalating an alert
When an alert is triggered, it begins escalating to the on-call responders immediately. For each escalation rule in the project's escalation policy, the designated on-call responders receive one email when the rule fires. You can respond to a page or stop alert escalations by updating the alert's status.
Escalating an incident
For incidents, paging on-call responders is optional for each individual incident. To begin escalating the incident, set the incident's escalation policy. For each escalation rule, the designated on-call responders receive one email when the rule fires. You can respond to a page or stop incident escalations by updating the incident's status or, if applicable, unsetting the incident's escalation policy.
To avoid duplicate pages, incidents created from alerts do not support independent escalation. Instead, the status and escalation policy fields are synced between the alert and the incident.