81-distributed-tracing

81 - Distributed Tracing with OpenTelemetry

This project demonstrates how to add Distributed Tracing to a Python microscopic application using OpenTelemetry.

Tracing is essential in microservice architectures to monitor requests that bounce between multiple services. You can trace bottlenecks, see exact times taken for database operations, and visualize the call tree.

Concept:

• FastAPI Backend: Emits tracing information (Spans and Traces) to the console using the ConsoleSpanExporter.
• Streamlit Frontend: Triggers requests to test the tracing in action.

Required Setup

# Install dependencies
pip install -r requirements.txt

Running the Application

1. Start the Traced API Server:

   python main.py

   This starts the FastAPI server on port 8000.

2. Open a new terminal and run the Frontend Dashboard:

   streamlit run app.py

3. Interact and Observe:

   • Click the button on the Streamlit UI to trigger a request.
   • Look at the terminal where main.py is running. You will see JSON blocks describing the Span, identifying the duration of both the root request and the nested sub-traces (heavy_processing_task, db_query).