Guides2 min read

Analytics

How your services and responders are performing — what each section shows, how filters, ranges and timezones work, what every metric means, and how to export.

Analytics answers "how are we doing, and where should we improve?" for the incidents you can see. It replaces the 2.1.0 Analytics page: every number is now computed in the database over every matching incident (no sampling), each section loads on its own, and the page works on phones — there is no separate mobile Analytics screen any more; /m links open /analytics.

Sections

Section What it shows
Key metrics Eight tiles: Incidents, Open now, MTTA, MTTR, Ack SLA met, Resolve SLA met, Escalated, After hours. MTTA and MTTR show the median with the p90 beneath.
Incident volume Incidents created per hour, day, week or month, split by urgency.
Time to acknowledge / Time to resolve Median and p90 per bucket.
Services Reliability per service. Select a service to filter the whole page.
Responder load Assigned incidents by time of day in each responder's own timezone, plus scheduled on-call hours.
When incidents happen Day-of-week by hour heatmap in your timezone.
Noise Alerts per incident, auto-resolved share, the loudest services and recurring incident titles.
Follow-through Postmortems published for resolved high-urgency incidents, and open and overdue action items.

A section that fails shows its own error with a retry; the rest of the page keeps working.

Permissions

  • View needs the report.read capability (REPORT_READ).
  • Export CSV needs report.export (REPORT_EXPORT); the button is hidden without it and the export endpoint returns 403.
  • Everything is scoped to the incidents you may read: users who are not administrators see incidents from services they can access, plus incidents assigned to them or that they watch. The page says so in a notice above the sections.

Last updated for v2.2.0

Edit this page on GitHub