FastAPI OpenTelemetry instrumentation

Instrument a FastAPI service with OpenTelemetry using the opentelemetry-instrument wrapper. Incoming requests, route handlers, database queries, and outbound HTTP calls are traced with no changes to your application code.

Fastest path. Run this in your AI IDE and the Instrumentation Agent wires this service up for you.

/osuite-instrument fastapi

The agent installs the FastAPI instrumentor and wires the exporter for your service.

Prefer to do it by hand? Follow the manual steps below — they are the complete path.

Prerequisites

• A FastAPI app served by Uvicorn (or Gunicorn with Uvicorn workers).
• Python 3.8 or newer.
• Your Osuite ingest token and region.

Install and configure

1. Install the distro and OTLP exporter, then let bootstrap add the FastAPI and ASGI instrumentors.

   pip install opentelemetry-distro opentelemetry-exporter-otlp
   opentelemetry-bootstrap -a install

2. Set the exporter environment variables.

   export OTEL_SERVICE_NAME="<service-name>"
   export OTEL_RESOURCE_ATTRIBUTES="service.environment=production"
   export OTEL_EXPORTER_OTLP_ENDPOINT="https://ingest.<region>.osuite.io:443"
   export OTEL_EXPORTER_OTLP_HEADERS="x-osuite-ingest-token=<your-ingest-token>"
   export OTEL_LOGS_EXPORTER="otlp"
   export OTEL_PYTHON_LOGGING_AUTO_INSTRUMENTATION_ENABLED="true"

Run FastAPI with instrumentation

These examples assume the app object is app in app/main.py.

Bare host

Run Uvicorn through the wrapper. Do not pass --reload (see gotchas):

opentelemetry-instrument uvicorn app.main:app --host 0.0.0.0 --port 8000

Docker

FROM python:3.12-slim

WORKDIR /app
COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt opentelemetry-distro opentelemetry-exporter-otlp \
    && opentelemetry-bootstrap -a install
COPY . .

ENV OTEL_LOGS_EXPORTER="otlp"
ENV OTEL_PYTHON_LOGGING_AUTO_INSTRUMENTATION_ENABLED="true"

CMD ["opentelemetry-instrument", "uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8000"]

Pass the exporter variables at runtime:

docker run \
  -e OTEL_SERVICE_NAME="<service-name>" \
  -e OTEL_RESOURCE_ATTRIBUTES="service.environment=production" \
  -e OTEL_EXPORTER_OTLP_ENDPOINT="https://ingest.<region>.osuite.io:443" \
  -e OTEL_EXPORTER_OTLP_HEADERS="x-osuite-ingest-token=<your-ingest-token>" \
  myimage

Kubernetes

apiVersion: apps/v1
kind: Deployment
metadata:
  name: fastapi-app
spec:
  template:
    spec:
      containers:
        - name: web
          image: myimage
          command: ["opentelemetry-instrument", "uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8000"]
          env:
            - name: OTEL_SERVICE_NAME
              value: "<service-name>"
            - name: OTEL_RESOURCE_ATTRIBUTES
              value: "service.environment=production"
            - name: OTEL_EXPORTER_OTLP_ENDPOINT
              value: "https://ingest.<region>.osuite.io:443"
            - name: OTEL_EXPORTER_OTLP_HEADERS
              value: "x-osuite-ingest-token=<your-ingest-token>"
            - name: OTEL_LOGS_EXPORTER
              value: "otlp"
            - name: OTEL_PYTHON_LOGGING_AUTO_INSTRUMENTATION_ENABLED
              value: "true"

Framework gotchas

Do not use --reload under the wrapper

Uvicorn’s --reload starts a file-watching supervisor that forks a worker process. Under opentelemetry-instrument the instrumented process is the supervisor, not the worker serving requests, so no traces appear. Run without --reload for any instrumented run.

Multiple workers fork after startup

Running uvicorn --workers N, or Gunicorn with uvicorn.workers.UvicornWorker, forks worker processes from the master. The exporter’s background thread does not survive the fork, so initialize the SDK per worker with a post_fork hook. See Gunicorn and uWSGI in the troubleshooting guide. A single Uvicorn worker under the wrapper needs no hook.

Verify in Osuite

What you should see

Send a request to any route. The service appears in APM within a minute with a trace per request, showing the route handler, database queries, and outbound calls as nested spans. Application logs appear in the Logs explorer with the matching trace ID.

Troubleshooting

No data in Osuite after a few minutes? Work through these checks.

• Endpoint — confirm the exporter targets exactly ingest.<region>.osuite.io:443 for your region, over TLS.
• Token — confirm the ingest header carries a valid <your-ingest-token> and has not been rotated.
• Pipeline — confirm the signal you expect (traces, logs, or metrics) is wired into an active pipeline with the otlp/osuite exporter attached.
• Export errors — check the application or Collector logs for OTLP export failures (auth, DNS, TLS, connection refused).
• Network — confirm the host has outbound access to the ingest endpoint on port 443.
• Timing — allow up to a minute for the first data to appear before assuming a failure.

Still stuck? Ask the Investigation Agent or contact support.

Next steps

Custom instrumentation Add manual spans and metrics for your business logic.

Explore traces Open the trace explorer and service map in Osuite.

Correlate logs with traces Link application logs to the traces that produced them.