Alerts
Alerts is currently in beta and may not be available for every account. Available conditions, management actions, and notification channels may vary depending on your role, plan, and enabled product areas.
The Alerts view monitors application telemetry and notifies your team when a rule's condition is met. Use alerts for conditions that should be detected without someone continuously watching the dashboard.
Developers and administrators can create and manage rules. Other members can inspect the rules and history available to them.
Choose a condition
A rule monitors one of these conditions:
| Condition | What it evaluates |
|---|---|
| Log count | Logs at selected levels within a time window. |
| New issue | A new issue, crash, or either type. |
| Issue event count | Issue or crash occurrences within a time window. |
| Devices affected | The number or percentage of active devices affected by issues or crashes. |
| Network 2xx, 3xx, 4xx, or 5xx | The corresponding HTTP response-class rate. |
| Network failed requests | The failed-request rate. |
| Network average duration | Mean request duration. |
| Network requests per minute | Request volume per minute. |
Network conditions are available only when the Network product area is enabled.
Configure thresholds
For conditions that use a threshold, choose:
- Static value to compare the current count or measurement with a fixed value;
- Percent to compare a supported rate or affected-device percentage with a fixed percentage;
- Percent change to compare the current window with the previous equal-length window;
- Above or Below to define which side of a static or percent threshold triggers the alert;
- a Critical threshold, and optionally a Warning threshold;
- an optional Recovery threshold to define when the alert returns to normal;
- a supported time window, from short incident windows to multi-day issue windows.
Positive percent-change values detect increases and negative values detect decreases. The available threshold types and windows depend on the selected condition.
Limit the scope
A broad rule can create noise. Depending on its condition, a rule can be limited by:
- log levels and tags;
- application versions;
- device or session ID;
- issue type: issues, crashes, or both;
- network domain and normalized endpoint.
Start broad enough to catch the behavior, then narrow the scope if unrelated traffic repeatedly triggers the rule.
Choose notification channels
Every rule requires at least one channel:
- Email sends notifications to the configured addresses.
- Webhook sends to an HTTPS endpoint you provide.
- Slack uses a Slack Incoming Webhook URL.
- In-app notifies team members through the dashboard inbox.
Treat webhook URLs as credentials. Limit access to them and rotate them if they are exposed.
Investigate a rule
Open a rule to review its current value, thresholds, severity, trigger count, last trigger time, scope, and notification channels. You can also pause or activate the rule without deleting its configuration.
The History table records trigger instances and notification delivery status. From history you can inspect an event, assign it, resolve it, or snooze it for 1 hour, 4 hours, 24 hours, or 7 days when those actions are available.
Deleting a rule permanently removes the rule and its history.