Incident Management
Learn how to effectively communicate service disruptions to your customers through incidents.
Creating an Incident
When a service disruption occurs, navigate to Incidents and click Report Incident. You'll need to provide:
- Title - A brief description of the issue (e.g., "API response times elevated")
- Status - The current state of your investigation
- Impact - How severely the issue affects users
- Message - Detailed information about what's happening and what you're doing
- Affected Components - Which services are impacted
When you create an incident, subscribers are automatically notified based on their notification preferences.
Incident Statuses
Incidents follow a structured workflow with four statuses:
You're aware of an issue and actively looking into it. This is the initial status for most incidents.
You've found the root cause and are working on a fix. Customers know you understand the problem.
A fix has been implemented and you're watching to ensure the issue doesn't recur.
The incident is fully resolved. Affected components return to their normal status.
Status Transitions
Statuses can only move forward in the workflow (Investigating → Identified → Monitoring → Resolved). The only exception is reopening a resolved incident, which returns it to Investigating.
Impact Levels
Impact levels help customers understand the severity of an incident:
| Level | Description | Example |
|---|---|---|
| None | Informational only, no user impact | Planned infrastructure changes |
| Minor | Some users may experience issues | Elevated latency, slow responses |
| Major | Significant functionality unavailable | API errors, feature outages |
| Critical | Complete service unavailability | Full outage, data unavailable |
Adding Timeline Updates
As you work on resolving an incident, add updates to keep customers informed. Each update includes:
- New Status - Optionally change the incident status
- Message - What's changed since the last update
Updates are displayed chronologically on your public status page, creating a timeline of the incident. Subscribers receive notifications for each update.
💡 Best Practice
Post updates every 30-60 minutes during active incidents, even if there's no new information. A simple "We're still investigating" reassures customers you're actively working on the issue.
Incident Templates
For common incident types, you can create templates to speed up incident creation. Templates pre-fill the title, impact level, message, and affected components.
When creating a new incident, select a template from the dropdown to apply it. You can still modify any fields before submitting.
Affected Components
Link incidents to the components they affect. This helps customers quickly see which services are impacted and allows subscribers to receive notifications only for components they care about.
You can add or remove affected components at any time during an incident by editing the incident details.