Distributed Tracing with Fastify, OpenTelemetry, Tempo, and Grafana

This project demonstrates distributed tracing across two Fastify applications (frontend and backend), utilizing OpenTelemetry for tracing, Tempo as a trace backend, and Grafana for trace visualization. The frontend service sends requests to the backend service, which handles random number generation and stores the numbers in a PostgreSQL database.

Features

• Distributed Tracing: Using OpenTelemetry with W3C Trace Context Propagator.
• Tempo Backend: Stores traces for distributed systems.
• Grafana: Visualize and explore traces using Tempo as a data source.
• PostgreSQL: Backend stores random numbers in a database.
• Docker & Docker Compose: Easily spin up the entire stack locally.
• Automated Testing: A script to send and fetch random numbers to test the system.

Prerequisites

Before running the project, ensure you have the following installed:

• Docker: Install Docker
• Docker Compose: Install Docker Compose
• jq: A command-line JSON processor (for viewing trace results). Install it via brew install jq (macOS) or apt-get install jq (Ubuntu).

Setup and Running Locally

make up

This will build and run the following services:

• Frontend (HTTP): http://localhost:3001
• Backend (HTTP): http://localhost:3002
• Grafana: http://localhost:3000 | Username: admin Password: admin

Visualize Traces in Grafana

After navigating to Grafana, log in with the credentials (admin / admin), and go to Explore. Select Tempo as the data source. You should be able to see the distributed traces from the frontend and backend.

Run the test script

To test the system by sending 100 random numbers to the write endpoint and retrieving them from the read endpoint, use the provided test.sh script.

Run the script as follows:

./test.sh

This will:

Send 100 requests to the write endpoint (triggering random number generation). Fetch the latest 100 numbers from the read endpoint and display them.

Stopping the Services

To stop all the services, use:

make down