Network Performance Monitoring is now generally available! Network Monitoring is now available!

Tracing Python Applications

For Python Django applications, note that tracing is disabled when your application is launched in DEBUG mode. Find more here.

Installation and Getting Started

If you already have a Datadog account you can find step-by-step instructions in our in-app guides for host-based and container-based set ups.

To begin tracing applications written in Python, first install and configure the Datadog Agent, see the additional documentation for tracing Docker applications or Kubernetes applications.

Next, install the Datadog Tracing library, ddtrace, using pip:

pip install ddtrace

Then to instrument your Python application use the included ddtrace-run command. To use it, prefix your Python entry-point command with ddtrace-run.

For example, if your application is started with python app.py then:

$ ddtrace-run python app.py

For more advanced usage, configuration, and fine-grain control, see Datadog’s API documentation.

Environment variable

When using ddtrace-run, the following environment variable options can be used:

Environment VariableDefaultDescription
DATADOG_TRACE_ENABLEDtrueEnable web framework and library instrumentation. When false, your application code doesn’t generate any traces.
DATADOG_ENVnullSet an application’s environment e.g. prod, pre-prod, staging`. Learn more about how to setup your environment.
DATADOG_TRACE_DEBUGfalseEnable debug logging in the tracer. Note that this is not available with Django.
DATADOG_SERVICE_NAMEnullOverride the service name to be used for this program. The value is passed through when setting up middleware for web framework integrations (e.g. pylons, flask, django). For tracing without a web integration, prefer setting the service name in code.
DATADOG_PATCH_MODULESnoneOverride the modules patched for this program execution. It should follow this format: DATADOG_PATCH_MODULES=module:patch,module:patch....
DD_AGENT_HOSTlocalhostOverride the address of the trace Agent host that the default tracer attempts to submit traces to.
DATADOG_TRACE_AGENT_PORT8126Override the port that the default tracer submit traces to.
DATADOG_PRIORITY_SAMPLINGtrueEnable Priority Sampling.
DD_LOGS_INJECTIONfalseEnable connecting logs and traces Injection.
DD_TRACE_ANALYTICS_ENABLEDfalseEnable App Analytics globally for web integrations.
DD_INTEGRATION_ANALYTICS_ENABLEDfalseEnable App Analytics for a specific integration. Example: DD_BOTO_ANALYTICS_ENABLED=true .

Change Agent Hostname

Configure your application level tracers to submit traces to a custom Agent hostname. The Python Tracing Module automatically looks for and initializes with the ENV variables DD_AGENT_HOST and DD_TRACE_AGENT_PORT

import os
from ddtrace import tracer

tracer.configure(
    hostname=os.environ['DD_AGENT_HOST'],
    port=os.environ['DD_TRACE_AGENT_PORT'],
)

Compatibility

Python versions 2.7 and 3.4 and onwards are supported.

Integrations

Web Framework Compatibility

The ddtrace library includes support for a number of web frameworks, including:

FrameworkSupported VersionPyPi Datadog Documentation
aiohttp>= 1.2http://pypi.datadoghq.com/trace/docs/web_integrations.html#aiohttp
Bottle>= 0.11http://pypi.datadoghq.com/trace/docs/web_integrations.html#bottle
Django>= 1.8http://pypi.datadoghq.com/trace/docs/web_integrations.html#django
djangorestframework>= 3.4http://pypi.datadoghq.com/trace/docs/web_integrations.html#django
Falcon>= 1.0http://pypi.datadoghq.com/trace/docs/web_integrations.html#falcon
Flask>= 0.10http://pypi.datadoghq.com/trace/docs/web_integrations.html#flask
Molten>= 0.7.0http://pypi.datadoghq.com/trace/docs/web_integrations.html#molten
Pylons>= 0.9.6http://pypi.datadoghq.com/trace/docs/web_integrations.html#pylons
Pyramid>= 1.7http://pypi.datadoghq.com/trace/docs/web_integrations.html#pyramid
Tornado>= 4.0http://pypi.datadoghq.com/trace/docs/web_integrations.html#tornado

Datastore Compatibility

The ddtrace library includes support for the following data stores:

DatastoreSupported VersionPyPi Datadog Documentation
Cassandra>= 3.5http://pypi.datadoghq.com/trace/docs/db_integrations.html#cassandra
Elasticsearch>= 1.6http://pypi.datadoghq.com/trace/docs/db_integrations.html#elasticsearch
Flask Cache>= 0.12http://pypi.datadoghq.com/trace/docs/db_integrations.html#flask-cache
Memcached pylibmc>= 1.4http://pypi.datadoghq.com/trace/docs/db_integrations.html#pylibmc
Memcached pymemcache>= 1.3http://pypi.datadoghq.com/trace/docs/db_integrations.html#pymemcache
MongoDB Mongoengine>= 0.11http://pypi.datadoghq.com/trace/docs/db_integrations.html#mongoengine
MongoDB Pymongo>= 3.0http://pypi.datadoghq.com/trace/docs/db_integrations.html#pymongo
MySQL MySQL-python>= 1.2.3http://pypi.datadoghq.com/trace/docs/db_integrations.html#module-ddtrace.contrib.mysqldb
MySQL mysqlclient>= 1.3http://pypi.datadoghq.com/trace/docs/db_integrations.html#module-ddtrace.contrib.mysqldb
MySQL mysql-connector>= 2.1http://pypi.datadoghq.com/trace/docs/db_integrations.html#mysql-connector
Postgres aiopg>= 0.12.0http://pypi.datadoghq.com/trace/docs/db_integrations.html#aiopg
Postgres psycopg>= 2.4http://pypi.datadoghq.com/trace/docs/db_integrations.html#module-ddtrace.contrib.psycopg
Redis>= 2.6http://pypi.datadoghq.com/trace/docs/db_integrations.html#redis
Redis redis-py-cluster>= 1.3.5http://pypi.datadoghq.com/trace/docs/db_integrations.html#module-ddtrace.contrib.rediscluster
SQLAlchemy>= 1.0http://pypi.datadoghq.com/trace/docs/db_integrations.html#sqlalchemy
SQLite3Fully Supportedhttp://pypi.datadoghq.com/trace/docs/db_integrations.html#sqlite
Vertica>= 0.6http://pypi.datadoghq.com/trace/docs/db_integrations.html#vertica

Library Compatibility

The ddtrace library includes support for the following libraries:

LibrarySupported VersionPyPi Datadog Documentation
asyncioFully Supportedhttp://pypi.datadoghq.com/trace/docs/async_integrations.html#asyncio
gevent>= 1.0http://pypi.datadoghq.com/trace/docs/async_integrations.html#gevent
aiobotocore>= 0.2.3http://pypi.datadoghq.com/trace/docs/other_integrations.html#aiobotocore
Boto2>= 2.29.0http://pypi.datadoghq.com/trace/docs/other_integrations.html#boto2
Botocore>= 1.4.51http://pypi.datadoghq.com/trace/docs/other_integrations.html#botocore
Celery>= 4.0.2http://pypi.datadoghq.com/trace/docs/other_integrations.html#celery
FuturesFully Supportedhttp://pypi.datadoghq.com/trace/docs/other_integrations.html#futures
Grpc>= 1.8.0http://pypi.datadoghq.com/trace/docs/other_integrations.html#grpc
httplibFully Supportedhttp://pypi.datadoghq.com/trace/docs/other_integrations.html#httplib
Jinja2>= 2.7http://pypi.datadoghq.com/trace/docs/other_integrations.html#jinja2
Kombu>= 4.0http://pypi.datadoghq.com/trace/docs/other_integrations.html#kombu
Mako>= 0.1.0http://pypi.datadoghq.com/trace/docs/other_integrations.html#mako
Requests>= 2.08http://pypi.datadoghq.com/trace/docs/other_integrations.html#requests

Further Reading