Optima Docs

Appearance

Performance Alerts ​

Optima's intelligent alerting system helps you catch performance issues before they impact your users. Set up sophisticated alerts with customizable thresholds, multiple notification channels, and actionable insights.

🚨 Alert System Overview ​

The alert system continuously monitors your application's performance metrics and notifies you when thresholds are exceeded. It supports:

  • Real-time Monitoring: Continuous evaluation of performance metrics
  • Smart Thresholds: Static values or percentage-based comparisons
  • Multiple Channels: Email and Slack notifications
  • Professional Templates: Beautiful, actionable alert messages
  • Team Collaboration: Multi-recipient notifications

📊 Supported Metrics ​

You can create alerts for any of these Core Web Vitals and performance metrics:

Core Web Vitals ​

  • LCP (Largest Contentful Paint): Loading performance
  • FID (First Input Delay): Interactivity measurement
  • CLS (Cumulative Layout Shift): Visual stability
  • INP (Interaction to Next Paint): Responsiveness
  • TTFB (Time to First Byte): Server response time
  • FCP (First Contentful Paint): Initial rendering

Custom Metrics ​

  • Error Rate: JavaScript error frequency
  • Loading Time: Custom page loading metrics
  • Custom Events: Business-specific performance indicators

⚙️ Alert Configuration ​

Basic Alert Setup ​

javascript
// Example alert configuration
{
  name: "LCP Performance Alert",
  description: "Alert when LCP exceeds acceptable thresholds",
  metric_name: "lcp",
  aggregation_type: "p95",
  evaluation_interval: "5m",
  
  // Threshold configuration
  critical_threshold: 4000,    // 4 seconds
  warning_threshold: 2500,     // 2.5 seconds
  resolved_threshold: 2000,    // 2 seconds
  
  // Notification settings
  is_enabled: true
}

Aggregation Types ​

Choose how your metric data is aggregated:

  • P50 (Median): 50th percentile - typical user experience
  • P75: 75th percentile - above-average user experience
  • P95: 95th percentile - worst-case scenarios (recommended)
  • P99: 99th percentile - extreme outliers
  • AVG: Average value across all sessions
  • MAX: Maximum value observed
  • MIN: Minimum value observed

Evaluation Intervals ​

Control how often alerts are evaluated:

  • 1m: Every minute (for critical metrics)
  • 5m: Every 5 minutes (recommended for most alerts)
  • 15m: Every 15 minutes (for less critical metrics)
  • 30m: Every 30 minutes
  • 1h: Every hour
  • 24h: Daily evaluation

📧 Notification Channels ​

Email Notifications ​

Professional HTML emails with comprehensive performance insights:

javascript
// Email notification configuration
{
  action_type: "email",
  trigger_status: "critical",
  settings: {
    email_recipients: [
      "dev-team@company.com",
      "performance@company.com"
    ]
  }
}

Email Features:

  • Professional Design: Mobile-responsive HTML templates
  • Performance Context: Detailed metric information and thresholds
  • Actionable Insights: Specific optimization recommendations
  • Visual Indicators: Color-coded status and severity levels
  • Direct Links: Quick access to detailed analysis in dashboard

Slack Integration ​

Real-time notifications in your team channels:

javascript
// Slack notification configuration
{
  action_type: "slack",
  trigger_status: "critical",
  settings: {
    webhook_url: "https://hooks.slack.com/services/YOUR/SLACK/WEBHOOK",
    channel: "#performance-alerts",
    message_template: "🚨 {{alert_name}} is {{status}}: {{metric_name}} = {{current_value}}ms (threshold: {{threshold}}ms)"
  }
}

Slack Features:

  • Real-time Delivery: Instant notifications to team channels
  • Custom Templates: Personalized message formats
  • Rich Formatting: Structured messages with context
  • Team Mentions: Alert specific team members or groups

🎯 Alert Status Levels ​

Critical Alerts ​

  • Trigger: When metrics exceed critical thresholds
  • Impact: Significant user experience degradation
  • Response: Immediate attention required
  • Notifications: All configured channels activated

Warning Alerts ​

  • Trigger: When metrics exceed warning thresholds
  • Impact: Noticeable performance degradation
  • Response: Investigation recommended
  • Notifications: Email notifications (configurable)

Resolved Alerts ​

  • Trigger: When metrics return below resolved thresholds
  • Impact: Performance has improved
  • Response: Confirmation of fix effectiveness
  • Notifications: Status update to all channels

🔧 Advanced Alert Configuration ​

Percentage-Based Thresholds ​

Set alerts based on percentage changes from baseline:

javascript
{
  name: "LCP Regression Alert",
  threshold_type: "percentage",
  comparison_period: "7d",
  critical_percentage: 25,     // 25% worse than last week
  warning_percentage: 15,      // 15% worse than last week
  resolved_percentage: 5       // Within 5% of baseline
}

Multi-Condition Alerts ​

Create complex alerts with multiple conditions:

javascript
{
  name: "Performance Degradation Alert",
  conditions: [
    {
      metric: "lcp",
      aggregation: "p95",
      threshold: 3000
    },
    {
      metric: "fid",
      aggregation: "p95", 
      threshold: 100
    }
  ],
  logic: "AND"  // Both conditions must be met
}

Custom Evaluation Windows ​

Fine-tune how alerts evaluate your data:

javascript
{
  name: "Custom Window Alert",
  evaluation_interval: "5m",
  evaluation_window: "15m",    // Look at last 15 minutes of data
  minimum_sessions: 10         // Require at least 10 sessions
}

📈 Alert Management Dashboard ​

Alert Overview ​

  • Alert Status: Current state of all alerts (Critical, Warning, Resolved)
  • Recent Triggers: Latest alert activations with timestamps
  • Performance Impact: How alerts correlate with user experience
  • Team Activity: Who responded to alerts and when

Alert History ​

  • Trigger Timeline: When alerts fired and resolved
  • Duration Analysis: How long performance issues lasted
  • Pattern Recognition: Recurring alert patterns
  • Resolution Tracking: Time to resolution metrics

Alert Analytics ​

  • False Positive Rate: Alerts that didn't require action
  • Response Time: How quickly team responds to alerts
  • Resolution Effectiveness: Whether fixes resolved issues
  • Cost Analysis: Impact of performance issues on business metrics

🛠️ Implementation Examples ​

E-commerce Site Alert ​

javascript
// Critical for checkout performance
{
  name: "Checkout Page LCP Alert",
  metric_name: "lcp",
  aggregation_type: "p95",
  critical_threshold: 2500,
  evaluation_interval: "1m",
  
  // Filter to specific pages
  filters: {
    page_url: "/checkout",
    device_type: "mobile"
  },
  
  actions: [
    {
      action_type: "email",
      trigger_status: "critical",
      settings: {
        email_recipients: ["ecommerce-team@company.com"]
      }
    },
    {
      action_type: "slack",
      trigger_status: "critical",
      settings: {
        webhook_url: "https://hooks.slack.com/...",
        channel: "#ecommerce-alerts"
      }
    }
  ]
}

SaaS Application Alert ​

javascript
// Monitor dashboard performance
{
  name: "Dashboard Performance Alert",
  metric_name: "inp",
  aggregation_type: "p95",
  critical_threshold: 200,
  warning_threshold: 100,
  evaluation_interval: "5m",
  
  // Business hours only
  schedule: {
    timezone: "America/New_York",
    active_hours: "09:00-17:00",
    active_days: ["monday", "tuesday", "wednesday", "thursday", "friday"]
  }
}

Mobile-First Alert ​

javascript
// Focus on mobile performance
{
  name: "Mobile LCP Alert",
  metric_name: "lcp",
  aggregation_type: "p95",
  critical_threshold: 3000,
  evaluation_interval: "5m",
  
  // Mobile devices only
  filters: {
    device_type: "mobile",
    connection_type: ["3g", "4g"]
  },
  
  // Custom message template
  actions: [{
    action_type: "slack",
    trigger_status: "critical",
    settings: {
      message_template: "📱 Mobile performance issue: LCP {{current_value}}ms on {{device_type}} ({{connection_type}})"
    }
  }]
}

🔍 Alert Troubleshooting ​

Common Issues ​

Alerts Not Triggering:

  • Verify evaluation interval and thresholds
  • Check minimum session requirements
  • Ensure sufficient data volume
  • Confirm alert is enabled

Too Many False Positives:

  • Adjust threshold values
  • Increase evaluation window
  • Add minimum session requirements
  • Use percentage-based thresholds

Missing Notifications:

  • Verify email addresses and Slack webhooks
  • Check spam folders for email alerts
  • Confirm Slack channel permissions
  • Test notification channels manually

Debug Mode ​

Enable detailed logging for alert troubleshooting:

javascript
// Add debug configuration
{
  name: "Debug Alert",
  debug: true,  // Enable detailed logging
  dry_run: true // Test without sending notifications
}

📊 Best Practices ​

Threshold Selection ​

  • Start Conservative: Begin with looser thresholds and tighten based on experience
  • Use P95: Recommended for catching issues without too much noise
  • Consider Context: Different pages may need different thresholds
  • Monitor Trends: Adjust thresholds based on performance improvements

Alert Hygiene ​

  • Regular Review: Audit alerts monthly for relevance
  • Team Alignment: Ensure alerts match team priorities
  • Documentation: Document what each alert monitors and why
  • Response Plans: Define clear procedures for each alert type

Notification Management ​

  • Right People: Send alerts to those who can act on them
  • Appropriate Channels: Use Slack for immediate issues, email for summaries
  • Escalation Paths: Define who to contact if primary responders are unavailable
  • Quiet Hours: Consider team schedules and time zones

🚀 Advanced Features ​

Alert Correlation ​

  • Multi-Metric Analysis: Correlate alerts across different metrics
  • Business Impact: Connect performance alerts to business metrics
  • Root Cause Analysis: Identify common causes of performance issues
  • Predictive Alerting: Alert on trends before thresholds are breached

Integration Capabilities ​

  • CI/CD Integration: Alert on deployment performance impact
  • Monitoring Tools: Connect with existing monitoring infrastructure
  • Incident Management: Integration with PagerDuty, Opsgenie, etc.
  • Business Intelligence: Export alert data for analysis

Custom Dashboards ​

  • Alert Widgets: Add alert status to custom dashboards
  • Performance Correlation: Show alerts alongside performance data
  • Team Dashboards: Create role-specific alert views
  • Executive Reporting: High-level alert summaries for leadership

Getting Started with Alerts ​

  1. Create Your First Alert - Basic alert setup
  2. Configure Notifications - Email and Slack setup
  3. Monitor Alert Performance - Track alert effectiveness
  4. Optimize Thresholds - Fine-tune your alerts

Ready to set up alerts? Visit your Optima dashboard to get started.

Copyright © 2025 Optima