Response

Incident Management

Track, manage, and resolve incidents created by failed healthchecks, missed heartbeats, and degraded monitors.

Overview

Uptimy tracks every outage as an incident, from detection to resolution. Incidents open automatically when a monitor meets its incident conditions: a healthcheck goes down, a heartbeat is missed, or a monitor set to alert when degraded becomes degraded. You can also report an incident by hand when your team finds an issue another way. Every incident has a timeline and records when it started and was resolved. An automatic incident also links to the monitor it is about.

Incidents connect monitors, workflows and status pages. When an automatic incident opens, matching workflows run your steps to notify responders, route alerts and escalate. Workflows run from monitor signals only: reporting an incident by hand, or changing an incident's status by hand, runs no workflows. To tell your users, publish the incident to your status pages. Its status changes and public updates then show there and reach the page's subscribers.

What Creates an Incident

An incident opens automatically from any of these sources:

  • Healthcheck failures — HTTP, gRPC, GraphQL, MCP, TCP, DNS, SSL, ICMP, or Playwright checks confirmed down
  • Missed heartbeats — A scheduled job, cron task, or background process does not ping within the expected window
  • Degraded monitors — Some regions fail or responses are slow, on a monitor set to alert when degraded
  • Uptimy Agent — A monitor an Uptimy Agent runs goes down, or a connected agent goes offline
  • Railway deploys — A deploy of a connected Railway service crashes or fails
  • Render services — A deploy of an imported Render service fails, or the service fails
  • Vercel deploys — A production deploy of an imported Vercel project fails
  • Dependencies — A third-party service you follow goes down

Automatic Incidents Trigger Workflows

An automatic incident is the handoff point between detection and response. Once it opens, workflows run your notification and escalation steps. Incidents you report by hand, and status changes you make by hand, run no workflows.

Incident Lifecycle

Every incident follows a lifecycle with six statuses. Each status change you make is recorded with a timestamp and who made it.

StatusDescription
CreatedThe incident is open. The app shows this step as Detected. For an automatic incident, this is when your workflows run.
AcknowledgedSomeone on your team has seen it. Posting the first update also acknowledges the incident. The time until the first status change feeds the MTTA metric.
InvestigatingThe team is actively investigating the root cause. Status updates can be posted to keep stakeholders informed.
IdentifiedThe root cause has been identified and the team is working on a fix. This signals progress to stakeholders.
MonitoringA fix has been deployed and the team is watching to confirm the service recovers and stays stable.
ResolvedThe issue is over. An automatic incident resolves on its own when its monitor recovers, and you can't resolve or reopen it by hand. You resolve an incident you reported with Resolve, and can Reopen it.

Automatic Incident Creation

When a monitor goes down or a heartbeat is missed, Uptimy creates an incident if one is not already open for that monitor. If one is open, the new event is added to its timeline instead of creating a duplicate. If the monitor fails again within 5 minutes of recovering, the same incident reopens.

Incident Timeline

Every incident keeps a timeline of what happened, from detection to resolution. Uptimy adds system events on its own, and your team adds updates.

  • Monitor events — The incident opening with the failure message, repeat failures, reopening, and recovery
  • Alerts sent — How many workflows the incident triggered
  • Status pages — When the incident was published to or removed from a status page
  • Status changes — Each status set by hand, with who set it
  • Updates — Messages your team posts, marked Public or Internal

Status Page Integration

An incident shows on a status page only once you publish it there. Until then it is visible only to your team in the dashboard, so you can triage without exposing anything to your users. Workflows for an automatic incident run either way.

  • Publish — Pick pages under Publish to status pages when you report an incident, or later with Edit details. Selected pages make it public; leave it empty to keep it internal.
  • Public or Internal updates — Each update you post is Public or Internal. Public updates show on the pages the incident is published to; internal ones stay with your team.
  • Subscribers — Subscribers of those pages hear when the incident is published, when its status changes, and when you post a public update. See Status Pages.

Communicate Early and Often

Post an initial status update within 5 minutes of detection, even if you do not yet know the root cause. Users appreciate knowing that you are aware of the issue. Follow up with updates every 15 to 30 minutes until the incident is resolved.

Incident Metrics

The Incidents page shows your workspace's incident numbers at the top.

MetricDescription
Total, Active and ResolvedHow many incidents the workspace has had, how many are still open, and how many are resolved.
MTTA (Mean Time to Acknowledge)Average time from the start of an incident to its first status change by your team, over the incidents someone acknowledged. Lower MTTA means your team reacts to incidents faster.
MTTR (Mean Time to Recover)Average time from the start of an incident to its resolution, over resolved incidents.

Post-Mortem

An incident's customer-facing summary is shown on the status pages it is published to. Once the incident is resolved, the app shows that summary as the Post-mortem. Use Edit details to write what happened, the impact and the fix. The timeline stays with the incident for review.

Blameless Post-Mortems

Focus on systemic improvements rather than individual blame. The incident timeline gives you an objective record of what happened and when.

Reporting Incidents Manually

Most incidents open automatically, but you can also report one from the Incidents page. Click Report incident and fill in:

  • Incident title — A short name for the issue
  • Severity — Critical, High, Medium or Low (High by default)
  • What's happening? — The impact, what you know so far, and what you are doing about it
  • Publish to status pages — The pages to show it on. Leave it empty to keep the incident internal.

A reported incident has the same lifecycle, timeline, updates and status page publishing as an automatic one. It runs no workflows. Subscribers of the pages you publish it to are notified.

Duplicate Prevention

Before creating a manual incident, check the active incidents list to ensure a related incident does not already exist. If a related incident is open, add your findings as a status update to the existing incident instead of creating a new one.