Alerts & Notifications

Stay informed about issues in your application with ActiveRabbit's flexible alerting system. Get notified through Slack, email, PagerDuty, or webhooks.

Alert Types

Error Alerts

• New Error - First occurrence of an error type
• Regression - Previously resolved error reoccurs
• Spike - Error count exceeds threshold
• High Impact - Error affecting many users

Performance Alerts

• Slow Response - P95 latency exceeds threshold
• Apdex Drop - User satisfaction score drops
• Throughput Change - Traffic anomaly detected
• Database Issue - Slow queries or N+1 detected

Notification Channels

Slack Integration

# 1. Install the Slack app from the dashboard
#    Settings → Integrations → Slack → Install

# 2. Configure channels for different alert types
# 
# #errors-critical   → New errors, regressions (P1)
# #errors-warnings   → Error spikes, high impact (P2)
# #performance       → Latency, Apdex alerts

Slack notifications include:

• Error summary with stack trace preview
• Quick link to full error details
• Action buttons: Resolve, Ignore, Assign
• AI-generated fix suggestion (when available)

Email Notifications

# Configure email alerts in the dashboard
# Settings → Notifications → Email
#
# Options:
# - Immediate: Send as errors occur
# - Digest: Daily/weekly summary
# - Threshold: Only when count exceeds N

PagerDuty Integration

# Connect PagerDuty for on-call alerting
# Settings → Integrations → PagerDuty
#
# Configure:
# - Service key
# - Severity mapping (Error level → PD severity)
# - Auto-resolve when error is resolved

Webhooks

# Send alerts to any HTTP endpoint
# Settings → Integrations → Webhooks

# Webhook payload example:
{
  "event": "error.new",
  "timestamp": "2024-01-15T10:30:00Z",
  "error": {
    "id": "err_abc123",
    "type": "NoMethodError",
    "message": "undefined method 'name' for nil",
    "count": 1,
    "users_affected": 1,
    "url": "https://app.activerabbit.ai/errors/err_abc123"
  },
  "project": {
    "id": "proj_xyz",
    "name": "MyApp"
  }
}

Alert Rules

Creating Custom Rules

# In the dashboard: Alerts → New Rule

# Example: Alert when checkout errors spike
# 
# Conditions:
#   - Tag: feature = "checkout"
#   - Error count > 10 in 5 minutes
#   - Environment: production
#
# Actions:
#   - Send to #checkout-team Slack channel
#   - Send to PagerDuty (High urgency)
#   - Send email to [email protected]

Rule Conditions

Programmatic Alerts

# Trigger alerts from code
ActiveRabbit.alert(
  title: "Payment processing failed",
  message: "Stripe webhook returned error: #{error.message}",
  severity: :critical,
  tags: {
    feature: 'payments',
    payment_id: payment.id
  }
)

# With custom notification routing
ActiveRabbit.alert(
  title: "Daily limit exceeded",
  severity: :warning,
  notify: ['#billing-team', '[email protected]']
)

Alert Fatigue Prevention

Deduplication

# Alerts are deduplicated by default
# Same error won't trigger multiple alerts within cooldown period

# Configure cooldown (in dashboard):
# - Default: 1 hour
# - Per rule: 5 minutes to 24 hours

Quiet Hours

# Set quiet hours for non-critical alerts
# Settings → Notifications → Quiet Hours
#
# During quiet hours:
# - Critical alerts: Always sent
# - High: Sent after 30 min if unresolved
# - Medium/Low: Batched for morning digest

Smart Batching

# Group related alerts together
# 
# Instead of 50 separate alerts:
# "🔴 Error spike detected
#  
#  50 new errors in the last 5 minutes:
#  - NoMethodError (23)
#  - ActiveRecord::RecordNotFound (15)
#  - Timeout::Error (12)
#  
#  Common factor: All from /api/v1/orders endpoint"

On-Call Schedules

# Integrate with your on-call rotation
# Settings → On-Call → Configure Schedule
#
# Supports:
# - PagerDuty schedules
# - Opsgenie
# - Custom webhook with @oncall mention

Alert History

View and analyze past alerts:

• Timeline view - See all alerts chronologically
• Correlation - Link alerts to deployments
• Statistics - MTTR, alert frequency, resolution rates
• Export - Download for analysis