Python SDK

Add Nolax to a Flask, FastAPI or Django app with a few lines. Install only the extras you need.

Package: nolax-sdk 1.1

Ingest contract: v1.1

Supported frameworks

Framework	Status
Flask	Supported
FastAPI	Supported
Django	Supported
Starlette (bare)	Supported

Install


pip install "nolax-sdk[flask]"

Flask example

python

import os
from flask import Flask
from nolax import NolaxFlask

app = Flask(__name__)
nolax = NolaxFlask(api_key=os.environ['NOLAX_API_KEY'], environment='production')
nolax.init_app(app)

FastAPI example

python

from fastapi import FastAPI
from nolax import NolaxFastAPI

app = FastAPI()
app.add_middleware(NolaxFastAPI, api_key=os.environ['NOLAX_API_KEY'], environment='production')

Warning

Django example

Add the middleware to settings.py and set the key and environment. Every other option is a setting named NOLAX_ plus the option name in capitals.

python

# settings.py
MIDDLEWARE = [
  "django.middleware.security.SecurityMiddleware",
  "django.contrib.sessions.middleware.SessionMiddleware",
  # ... other middleware ...
  "nolax.middleware.NolaxDjango",
]

NOLAX_API_KEY = os.environ.get("NOLAX_API_KEY")
NOLAX_ENVIRONMENT = "production"

# Optional
NOLAX_BATCH_SIZE = 100
NOLAX_FLUSH_INTERVAL_MS = 2000

The Django middleware works from SDK 1.1. Earlier versions could not start it.

Config options

Flask and FastAPI take these as keyword arguments. Django reads them from settings.

Option	Type	Default	Notes
api_key	str	required	Project API key
environment	str	required	development, staging or production
endpoint	str	https://server.nolax.net/ingest	Ingest URL, for local testing
batch_size	int	100	Events that trigger a flush
flush_interval_ms	int	2000	Milliseconds between timed flushes
max_queue_size	int	10000	Events held in memory. The oldest are dropped first
max_in_flight	int	4	Accepted for parity with the other SDKs. Python sends from one thread
release	str	None	Release identifier, up to 64 characters
service	str	None	Service name, up to 64 characters
consumer	callable	None	Takes the request, returns a consumer identifier
capture_ip	bool	True	Send the client IP address
capture_user_agent	bool	True	Send the User-Agent header
ip_header	str	None	Header that holds the client IP behind a proxy, for example x-forwarded-for
ignore_paths	list	[]	Strings or compiled patterns. Matching requests are never recorded
ignore_methods	list	[]	Methods that are never recorded
before_send	callable	None	Change an event, or return None to drop it
dry_run	bool	False	Log each batch with the key masked and send nothing. Also NOLAX_DRY_RUN=1
debug	bool	False	Verbose logging through the nolax logger
on_error	callable	None	Called when a batch is dropped

Flushing on shutdown

Call flush() on the Flask or FastAPI integration object to send queued events before the process exits. The SDK also flushes when the interpreter shuts down. It restarts its timer in worker processes created by fork, so Gunicorn and uWSGI workers keep sending.

Troubleshooting

No events in FastAPI: call app.add_middleware(...) before the first request arrives.
No events in Django: check that NOLAX_API_KEY, NOLAX_ENVIRONMENT and the middleware are in settings.py.
payloadSize is 0 when the response has no Content-Length header.
ip shows your proxy: set ip_header to x-forwarded-for, and make sure only your proxy can set that header.