Spike Detection Alerts
Automatically detect when new issues appear at 2x or more of their hourly baseline and get alerted through your project's integrations.
AbrovaTrace monitors how fast new issues appear and automatically detects when that rate spikes above normal levels. When a spike is detected, an alert goes to your project's active integrations so you can investigate before it affects more users.
Spike Detection#
Automatically detects when the number of new issues in the last hour reaches 200% (2x) or more of the hourly baseline and sends an alert through your project's integrations — so you can investigate before the spike affects more users.
How It Works
- Continuous Monitoring — Every 15 minutes, AbrovaTrace counts new issues (errors seen for the first time) per project and platform.
- Baseline Comparison — The number of new issues in the last hour is compared with the hourly average of the 23 hours before it.
- Spike Detection — If the last hour has at least 3 new issues and reaches 200% (2x) or more of the baseline, a spike is detected. With no baseline at all, more than 5 new issues in the last hour counts as a spike.
- Alert Sent — A notification is sent to every active integration of the project (Email, Slack, Discord, Telegram, Webhook, GitHub). If you have no integrations set up, no alert is sent.
Visual Example
New issues suddenly jump to 5 in the last hour when the baseline was 1 per hour:
Alert Email Preview (Email integration)
Subject: [AbrovaTrace] New Error: 🚨 Error Spike Detected: 500% increase in error rate
🚨 Error Spike Detected: 500% increase in error rate
Project: My Flutter App
5
Occurrences (new issues, last hour)
~1
Baseline (avg/hour)
Error rate increased from ~1/hour to 5/hour
Configuration
Spike detection runs automatically for every project. There is nothing to switch on. Current settings:
| Option | Type | Required | Description |
|---|---|---|---|
| Check Interval | - | - | Every 15 minutes |
| Spike Threshold | - | - | 200% (2x) of baseline, at least 3 new issues |
| Comparison Window | - | - | Last 1 hour |
| Baseline Window | - | - | Hourly average of the previous 23 hours |
| Alert Method | - | - | The project's active integrations |
Need your own threshold, window or channels? Create an Error spike rule under Trace → Alerts. Managing alert rules needs the admin or owner role; members can acknowledge and mute alerts — see Roles & Permissions.
Best Practices
When you receive a spike alert, investigate immediately. Check recent deployments, third-party services, or infrastructure issues.
Spike alerts go only to integrations — configure Email or a Webhook under Trace → Integrations. Email and webhooks work without a connection outside Iran.
Combine with Ignore Rules to filter out known errors and reduce false spike alerts.
Troubleshooting
No spike alert arrived?
- Check that the project has at least one integration and that it is switched on in Trace → Integrations
- Use "Send test" on the integration to confirm it delivers
- Spikes count new issues, not repeated events of an issue you already have — a flood of the same error does not count as a spike
- The last hour needs at least 3 new issues. A brand-new project has no baseline yet: then more than 5 new issues in the last hour count as a spike
- Integrations that deliver to Slack, Discord, Telegram or GitHub need those services to be reachable — see each integration's page