092 — Monitoring & Alerting {DevOps}

Purpose

Establish comprehensive observability across the application stack. Detect crashes, performance degradation, and downtime proactively through Firebase and Google Cloud monitoring tools. [EXPLICIT]

Physics — 3 Immutable Laws

1. Law of Observability: If it's not monitored, it's not in production. Every service (hosting, functions, firestore, auth) has active monitoring. [EXPLICIT]
2. Law of Alert Fatigue Prevention: Alerts fire only on actionable conditions. No informational alerts — only conditions requiring human intervention. [EXPLICIT]
3. Law of Mean Time to Detect: MTTD < 5 minutes for critical issues. Uptime checks run every 60 seconds. Crash reports arrive in real-time. [EXPLICIT]

Protocol

Phase 1 — Crash & Error Monitoring

1. Enable Firebase Crashlytics for web: import { getPerformance } from 'firebase/performance'. [EXPLICIT]
2. Configure global error boundary in React to report uncaught errors. [EXPLICIT]
3. Cloud Functions: structured logging with functions.logger — errors auto-surface in Cloud Logging. [EXPLICIT]
4. Set up Cloud Logging sink for error-level logs → alert channel. [EXPLICIT]

Phase 2 — Performance Monitoring

1. Initialize Firebase Performance Monitoring SDK in app entry point. [EXPLICIT]
2. Add custom traces for critical operations: perf.trace('checkout-flow'). [EXPLICIT]
3. Monitor Cloud Functions execution time via Cloud Monitoring metrics. [EXPLICIT]
4. Set performance budgets: page load < 3s, function execution < 10s. [EXPLICIT]

Phase 3 — Alerts & Uptime

1. Cloud Monitoring uptime check: HTTPS GET on production URL every 60s. [EXPLICIT]
2. Alert policies: error rate > 1% → email + Slack. Uptime check fails 2x → page. [EXPLICIT]
3. Firestore usage alert: reads > 50K/day or writes > 10K/day → email. [EXPLICIT]
4. Billing alert: projected spend > budget threshold → email + pause non-critical functions. [EXPLICIT]

I/O

Input	Output
Application errors/crashes	Crashlytics dashboard + alert
Page loads and custom traces	Performance Monitoring dashboard
Cloud Functions logs	Cloud Logging queries + alerts
Production URL	Uptime check status (up/down)

Quality Gates — 5 Checks

1. Crashlytics enabled — crash-free rate visible in Firebase Console. [EXPLICIT]
2. Uptime check active — 60-second interval on production URL. [EXPLICIT]
3. Alert channels configured — email + Slack/PagerDuty for critical alerts. [EXPLICIT]
4. Billing alerts set — threshold at 80% of monthly budget. [EXPLICIT]
5. Error rate dashboard exists — real-time error rate visible to team. [EXPLICIT]

Edge Cases

• Crash in error boundary: Implement fallback logging (beacon API) for catastrophic failures.
• Cold start noise: Exclude first invocation from function performance baselines.
• Alert storms: Set alert cooldown period (5 min) to prevent notification flooding.
• Third-party outages: Monitor external API dependencies separately from app health.

Self-Correction Triggers

• Crash-free rate drops below 99.5% → immediate investigation and hotfix.
• Uptime check fails → verify hosting, DNS, SSL certificate status.
• Performance regression detected → run Lighthouse audit (skill 096).
• Billing spike → audit Firestore queries (skill 100), check for infinite loops in functions.

Usage

Example invocations:

• "/monitoring-alerting" — Run the full monitoring alerting workflow
• "monitoring alerting on this project" — Apply to current context

Assumptions & Limits

• Assumes access to project artifacts (code, docs, configs) [EXPLICIT]
• Requires English-language output unless otherwise specified [EXPLICIT]
• Does not replace domain expert judgment for final decisions [EXPLICIT]