Skip to main content

OpenTracing support for Django applications

Project description

This package enables distributed tracing in Django projects via The OpenTracing Project. Once a production system contends with real concurrency or splits into many services, crucial (and formerly easy) tasks become difficult: user-facing latency optimization, root-cause analysis of backend errors, communication about distinct pieces of a now-distributed system, etc. Distributed tracing follows a request on its journey from inception to completion from mobile/browser all the way to the microservices.

As core services and libraries adopt OpenTracing, the application builder is no longer burdened with the task of adding basic tracing instrumentation to their own code. In this way, developers can build their applications with the tools they prefer and benefit from built-in tracing instrumentation. OpenTracing implementations exist for major distributed tracing systems and can be bound or swapped with a one-line configuration change.

If you want to learn more about the underlying python API, visit the python source code.

If you are migrating from the 0.x series, you may want to read the list of breaking changes.


Run the following command:

$ pip install signalfx-instrumentation-django

Setting up Tracing

In order to implement tracing in your system, add the following lines of code to your site’s file:

import django_opentracing

# OpenTracing settings

# if not included, defaults to True.
# has to come before OPENTRACING_TRACING setting because python...

# defaults to []
# only valid if OPENTRACING_TRACE_ALL == True

# Callable that returns an `opentracing.Tracer` implementation.

# Parameters for the callable (Depending on the tracer implementation chosen)
    'example-parameter-host': 'collector',

If you want to directly override the DjangoTracing used, you can use the following. This may cause import loops (See #10)

# some_opentracing_tracer can be any valid OpenTracing tracer implementation
OPENTRACING_TRACING = django_opentracing.DjangoTracing(some_opentracing_tracer)

Note: Valid request attributes to trace are listed [here]( When you trace an attribute, this means that created spans will have tags with the attribute name and the request’s value.

Tracing All Requests

In order to trace all requests, OPENTRACING_TRACE_ALL needs to be set to True (the default). If you want to trace any attributes for all requests, then add them to OPENTRACING_TRACED_ATTRIBUTES. For example, if you wanted to trace the path and method, then set OPENTRACING_TRACED_ATTRIBUTES = ['path', 'method'].

Tracing all requests uses the middleware django_opentracing.OpenTracingMiddleware, so add this to your file’s MIDDLEWARE_CLASSES at the top of the stack.

    ... # other middleware classes

Tracing Individual Requests

If you don’t want to trace all requests to your site, set OPENTRACING_TRACE_ALL to False. Then you can use function decorators to trace individual view functions. This can be done by adding the following lines of code to (or any other file that has url handler functions):

from django.conf import settings

tracing = settings.OPENTRACING_TRACING

def some_view_func(request):
    ... # do some stuff

This tracing method doesn’t use middleware, so there’s no need to add it to your file.

The optional arguments allow for tracing of request attributes. For example, if you want to trace metadata, you could pass in @tracing.trace('META') and request.META would be set as a tag on all spans for this view function.

Note: If OPENTRACING_TRACE_ALL is set to True, this decorator will be ignored, including any traced request attributes.

Accessing Spans Manually

In order to access the span for a request, we’ve provided an method DjangoTracing.get_span(request) that returns the span for the request, if it is exists and is not finished. This can be used to log important events to the span, set tags, or create child spans to trace non-RPC events.

Tracing an RPC

If you want to make an RPC and continue an existing trace, you can inject the current span into the RPC. For example, if making an http request, the following code will continue your trace across the wire:

def some_view_func(request):
    new_request = some_http_request
    current_span = tracing.get_span(request)
    text_carrier = {}
    opentracing_tracer.inject(span, opentracing.Format.TEXT_MAP, text_carrier)
    for k, v in text_carrier.items():
    ... # make request


Here is an example of a Django application that acts as both a client and server, with integrated OpenTracing tracers.

Breaking changes from 0.x

Starting with the 1.0 version, a few changes have taken place from previous versions:

  • DjangoTracer has been renamed to DjangoTracing, although DjangoTracer can be used still as a deprecated name. Likewise for OPENTRACING_TRACER being renamed to OPENTRACING_TRACING.

  • When using the middleware layer, OPENTRACING_TRACE_ALL defaults to True.

  • When no opentracing.Tracer is provided, DjangoTracing will rely on the global tracer.

Further Information

If you’re interested in learning more about the OpenTracing standard, please visit or join the mailing list. If you would like to implement OpenTracing in your project and need help, feel free to send us a note at

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

signalfx-instrumentation-django-1.2.0.tar.gz (25.1 kB view hashes)

Uploaded source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page