AutoSRE Platform

🌐 Live Demo: auto-sre.vercel.app

A production-like Site Reliability Engineering platform built to demonstrate observability, failure handling, and automated recovery in distributed systems.

🎯 What is AutoSRE?

AutoSRE simulates real-world production challenges that SRE teams face daily. It's a working demonstration of how to design, monitor, break, and heal microservices at scale.

Built to showcase:

• Instrumentation and observability (Prometheus metrics)
• Self-healing patterns (circuit breakers, retries)
• Chaos engineering (controlled failure injection)
• Incident response (automated recovery + postmortems)

✅ Current Features (Phase 1 Complete)

• Instrumented API Service - FastAPI with RED metrics (Rate, Errors, Duration)
• Prometheus Integration - Counter, Histogram, and Gauge metrics with proper labeling
• Health Checks - Liveness probes for container orchestration
• Docker Containerization - Reproducible deployments across environments
• Auto-generated API Docs - Interactive documentation at /docs

🚧 Roadmap

Phase 2: Observability Stack (In Progress)

• Prometheus auto-scraping with 15s intervals
• Grafana dashboards with SLO tracking
• Alert definitions based on SLI violations

Phase 3: Multi-Service Architecture

• Worker service (async job processing)
• Auth service (with circuit breaker pattern)
• PostgreSQL + Redis integration

Phase 4: Chaos Engineering

• Failure injection endpoints
• Latency injection
• Database connection failures
• Self-healing validation

Phase 5: Production Readiness

• Incident postmortems (3 documented scenarios)
• Runbook automation
• Load testing results
• Performance benchmarks

🚀 Quick Start

Prerequisites

• Docker Desktop installed
• Git

Run Locally

# Clone the repository
git clone https://github.com/Raynzler/Auto-SRE.git
cd Auto-SRE

# Start all services
docker-compose up --build

Access Services

• API Service: http://localhost:8000
• API Docs: http://localhost:8000/docs
• Metrics Endpoint: http://localhost:8000/metrics
• Prometheus: http://localhost:9090 (Phase 2)
• Grafana: http://localhost:3000 (Phase 2)

Test the API

# Health check
curl http://localhost:8000/health

# Create an order
curl -X POST http://localhost:8000/orders \
  -H "Content-Type: application/json" \
  -d '{"item": "laptop", "quantity": 1}'

# View metrics
curl http://localhost:8000/metrics

🛠 Technology Stack

Backend: Python, FastAPI, Pydantic
Observability: Prometheus, Grafana
Infrastructure: Docker, Docker Compose
Database: PostgreSQL (Phase 3)
Cache: Redis (Phase 3)
Proxy: Nginx (Phase 3)
Deployment: Vercel (landing page), Docker (services)

📊 Key Metrics Tracked

• Request Rate: Requests per second by endpoint and method
• Error Rate: Percentage of failed requests (4xx, 5xx)
• Latency: p50, p95, p99 response times
• Service Health: Binary indicator (up/down)
• Saturation: Resource usage (CPU, memory) (Phase 3)

🎓 Learning Outcomes

This project demonstrates:

• ✅ SRE fundamentals (SLIs, SLOs, error budgets)
• ✅ Instrumentation best practices (RED metrics)
• ✅ Container orchestration
• ✅ Failure mode analysis
• ✅ Automated recovery patterns
• ✅ Incident documentation

📝 Documentation

• Architecture Decisions (coming soon)
• Incident Postmortems (Phase 5)
• Runbook (Phase 5)

🤝 Contributing

This is a personal learning project, but feedback is welcome! Open an issue or PR if you spot improvements.

📧 Contact

Built by: Hamza Shaikh GitHub: @Raynzler
Project Status: 🟢 Active Development

---

This project is designed for portfolio demonstration and SRE role interviews. It mirrors production patterns used at companies like Google, Netflix, and Stripe.