Auto Instrumented FastAPI Application

This is a simple FastAPI application that is instrumented with OpenTelemetry using the OpenTelemetry Auto Instrumentation for Python. It listens on the port 8000, and generates telemetry data for each request.

Environment variables

A full reference of the Opentelemetry environment variables can be found here: opentelemetry/environment-variables

OTEL_SERVICE_NAME: The name of the service generating the telemetry data.

OTEL_LOGS_EXPORTER: The logs exporter to use, In this doc we're using the console exporter. If you have an opentelemetry collector set up, you can change this variable's value to console,otlp or just otlp. You will also have to set the OTEL_EXPORTER_OTLP_ENDPOINT variable to the endpoint of your opentelemetry collector.

OTEL_TRACES_EXPORTER: The traces exporter to use, In this doc we're using the console exporter. If you have an opentelemetry collector set up, you can change this variable's value to console,otlp or just otlp. You will also have to set the OTEL_EXPORTER_OTLP_ENDPOINT variable to the endpoint of your opentelemetry collector.

OTEL_METRICS_EXPORTER: The metrics exporter to use, In this doc we're using the console exporter. If you have an opentelemetry collector set up, you can change this variable's value to console,otlp or just otlp. You will also have to set the OTEL_EXPORTER_OTLP_ENDPOINT variable to the endpoint of your opentelemetry collector.

OTEL_PYTHON_LOGGING_AUTO_INSTRUMENTATION_ENABLED: This variable enables auto-instrumentation for the logging module.

Using the DockerHub repository

The image for this application is published on DockerHub, it can be found here: rutush10/otel-autoinstrumentation-fastapi-simple-app

To use the image with simple docker run command:

docker run
	-e OTEL_SERVICE_NAME=single-app-single-collector-fastapi-app \
	-e OTEL_LOGS_EXPORTER=console \
	-e OTEL_TRACES_EXPORTER=console \
	-e OTEL_METRICS_EXPORTER=console \
	-e OTEL_PYTHON_LOGGING_AUTO_INSTRUMENTATION_ENABLED=true \
	-p 8000:8000 \
	rutush10/otel-autoinstrumentation-fastapi-simple-app:latest

To use it within a docker-compose.yaml:

services:
  app:
    image: rutush10/otel-autoinstrumentation-fastapi-simple-app:latest
    ports:
      - "8000:8000"
    environment:
      - OTEL_SERVICE_NAME=otel-autoinstrumentation-fastapi-simple-app
      - OTEL_LOGS_EXPORTER=console
      - OTEL_TRACES_EXPORTER=console
      - OTEL_METRICS_EXPORTER=console
      - OTEL_PYTHON_LOGGING_AUTO_INSTRUMENTATION_ENABLED=true

To create a k8s deployment:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: simple-fastapi-app
  labels:
    app: simple-fastapi-app
spec:
  replicas: 1
  selector:
    matchLabels:
      app: simple-fastapi-app
  template:
    metadata:
      labels:
        app: simple-fastapi-app
    spec:
      containers:
      - name: simple-fastapi-app
        imagePullPolicy: Always
        image: rutush10/otel-autoinstrumentation-fastapi-simple-app:latest
        ports:
          - containerPort: 8000
        environment:
          -e OTEL_SERVICE_NAME=simple-fastapi-app
          -e OTEL_LOGS_EXPORTER=console
          -e OTEL_TRACES_EXPORTER=console
          -e OTEL_METRICS_EXPORTER=console
          -e OTEL_PYTHON_LOGGING_AUTO_INSTRUMENTATION_ENABLED=true

If you're working with this applicaiton within a k8s environment, don't forget to add a NodePort or a ClusterIP service:

apiVersion: v1
kind: Service
metadata:
  name: simple-fastapi-app
  labels:
    app: simple-fastapi-app
spec:
  type: NodePort
  selector:
    app: simple-fastapi-app
  ports:
    - protocol: TCP
      port: 8000
      targetPort: 8000
      nodePort: 30000

For a ClusterIP service, change the ports to

  ports:
    - protocol: TCP
      port: 8000
      targetPort: 8000

Testing the application locally

Build

make build-app TAG=0.0.1

Run

make run-app TAG=0.0.1

Stop

make stop-app

Make Requests

The postman directory contains the postman collection and the environment for testing the application locally. You have to simply import the collection and environment in your postman and run the requests.

Examples are provided within the requests for reference.

Auto Instrumenting FastAPI Applications

The opentelemetry-instrument command line tool can be used to automatically instrument a FastAPI application. In this case we define the application without any instrumentation code:

import uvicorn
from fastapi import FastAPI

app = FastAPI()

@app.get("/")
def health_check():
    return {"message": "Hello World"}

def main():
    uvicorn.run(app, host="0.0.0.0", port=8000)

if __name__ == "__main__":
    main()

Then we run application via the opentelemetry-instrument command line tool.

opentelemetry-instrument python main.py

You can see the run command in build/Dockerfile for this application.

Using this code as a reference for your own application

If you want to use this code as a reference for your own FastAPI application, you can copy the following to your own application:

1. The build directory, which contains the Dockerfile and the requirements.txt file. You can edit the requirements.txt file to add any additional dependencies that you might need.

2. The src directory, which contains the source code. You can add more files and folders as needed.

3. The Makefile, which contains the commands to build, run, publish, and stop the application. To publish the application to your own repository, edit the REPOSITORY, APP_IMAGE_NAME, and the DOCKER_REPO_NAME variables in the Makefile.