Alert Manager Overview

Intelligent alerting system with ML-based anomaly detection.

Features

Smart Alerting

• Automatic anomaly detection
• Baseline learning
• Trend prediction
• Noise reduction

Multiple Channels

• Email
• Slack
• Telegram
• Discord
• Webhooks
• SMS (Twilio)
• Voice calls (Twilio)

Alert Grouping

• Deduplicate similar alerts
• Group related incidents
• Smart notification throttling

Alert Types

Monitor Alerts

Triggered by monitor checks:

• Monitor down (failure threshold reached)
• Monitor recovered
• SSL certificate expiring
• Response time spike

Anomaly Alerts

ML-detected unusual patterns:

• Traffic spike
• Error rate increase
• Response time degradation
• Resource usage anomaly

Quota Alerts

Usage thresholds:

• RUM quota 80% used
• OTLP quota exceeded
• Top-up balance low

Custom Alerts

Based on custom metrics:

• Business KPI thresholds
• User-defined conditions
• Scheduled reports

Creating Alerts

Dashboard → Alerts → Create Alert

1. Select Trigger:

   • Monitor status
   • Anomaly detection
   • Quota threshold
   • Custom metric

2. Configure Conditions:

   • Threshold values
   • Time windows
   • Severity levels

3. Choose Channels:

   • Email, Slack, etc.
   • Multiple channels supported

4. Set Schedule:

   • 24/7 (default)
   • Business hours only
   • Custom schedule

Alert States

• Active: Condition met, notifications sent
• Acknowledged: Team notified, investigating
• Resolved: Issue fixed, auto-closed
• Muted: Temporarily silenced

Next Steps

• Anomaly Detection - ML alerting
• Channels - Configure notifications
• Thresholds - Fine-tune triggers