Load Testing Demo with Locust, Prometheus & Grafana

This project demonstrates how to set up a load testing and monitoring stack using:

• FastAPI → sample web app (our test subject)
• Locust → simulate user traffic
• Locust Exporter → expose metrics in Prometheus format
• Prometheus → collect & store metrics
• Grafana → visualize performance trends

I also wrote a detailed article explaining this setup step by step on Medium:
From Stress to Success: Load Testing Python Apps & Visualizing Performance

---

📂 Project Structure

server/
├── app.py              # FastAPI sample app
├── locustfile.py       # Locust test definition (user behavior)
├── prometheus.yml      # Prometheus scrape config
├── docker-compose.yml  # Combined setup (exporter, Prometheus, Grafana)
├── poetry.lock         # Dependency lock file (Poetry)
├── pyproject.toml      # Python project dependencies

---

🛠 Prerequisites

• Docker + Docker Compose
• Python 3.9+ (if running app/Locust locally)
• Poetry for dependency management

---

▶Getting Started

1. Move into the project folder

cd server

---

2. Run the FastAPI app

poetry install
poetry run uvicorn app:app --host 0.0.0.0 --port 8000

App available at → http://localhost:8000

---

3. Run Locust (load testing)

poetry run locust -f locustfile.py --host=http://localhost:8000

Locust UI → http://localhost:8089

---

4. Start Monitoring Stack (Exporter, Prometheus, Grafana)

docker compose up

Services available at:

• Locust Exporter → http://localhost:9646/metrics
• Prometheus → http://localhost:9090
• Grafana → http://localhost:3000

Default Grafana login:

username: admin
password: admin

---

📊 Visualizing Performance

In Grafana:

1. Add Prometheus datasource → http://prometheus:9090
2. Import or create dashboards with queries like:

• Requests per second:

  rate(locust_requests_total[1m])
  

• Avg response time:

  locust_response_time_seconds_avg
  

• 95th percentile latency:

  locust_response_time_seconds_percentile_95
  

• Failure rate:

  rate(locust_failures_total[5m])
  

---

What You’ll Learn

• How to simulate user traffic against a Python web app
• How to expose Locust metrics in Prometheus format
• How to scrape and visualize load test data over time
• How to spot performance bottlenecks before production

---

📜 License

MIT — feel free to fork and adapt.