Uptime + incidents via Better Stack; backend event analytics via PostHog.
This project demonstrates monitoring, alerting, and incident management using Better Stack through an intentional failure scenario.
A dedicated endpoint (/chaos) was designed to return intermittent 5xx responses. An uptime monitor detects these failures and automatically triggers an incident, which is communicated via a public status page.
The incident is acknowledged, updated, and resolved once the service is stabilized and monitoring confirms recovery. This mirrors a real-world incident lifecycle: detection → investigation → communication → resolution.
The status page also includes a completed scheduled maintenance window to demonstrate planned operational workflows.
Status page: https://status.lillacsanaky.dev/
The API emits explicit PostHog events to track endpoint usage, failure rate, and request latency. This complements uptime monitoring and incident alerts.
Environment variables:
- POSTHOG_API_KEY
- POSTHOG_HOST (EU: https://eu.i.posthog.com)
Tiny FastAPI service with:
GET /health→ always returns 200GET /chaos→ intermittently returns 500 (simulates a flaky dependency)
Used to demonstrate Better Stack monitoring, alerting, and incident lifecycle handling.