Opentelemetry auto instrumentation for django outbox pattern library

This library will help you to use opentelemetry traces and metrics on Django outbox pattern usage library.

Installation

pip install opentelemetry-instrumentation-django-outbox-pattern

How to use ?

You can use the DjangoOutboxPatternInstrumentor().instrument() for example in manage.py file.

#!/usr/bin/env python
"""Django's command-line utility for administrative tasks."""
import os
import sys
import typing

from opentelemetry_instrumentation_django_outbox_pattern import DjangoOutboxPatternInstrumentor

from opentelemetry import trace
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import BatchSpanProcessor
from opentelemetry.sdk.trace.export import ConsoleSpanExporter
from opentelemetry.trace.span import Span


def publisher_hook(span: Span, body: typing.Dict, headers: typing.Dict):
    # Custom code in your project here we can see span attributes and make custom logic with then.
    pass


def consumer_hook(span: Span, body: typing.Dict, headers: typing.Dict):
    # Custom code in your project here we can see span attributes and make custom logic with then.
    pass


provider = TracerProvider()
trace.set_tracer_provider(provider)
provider.add_span_processor(BatchSpanProcessor(ConsoleSpanExporter()))

def main():
    os.environ.setdefault("DJANGO_SETTINGS_MODULE", "application.settings")
    DjangoOutboxPatternInstrumentor().instrument(
        trace_provider=trace,
        publisher_hook=publisher_hook,
        consumer_hook=consumer_hook,
    )
    try:
        from django.core.management import execute_from_command_line
    except ImportError as exc:
        raise ImportError(
            "Couldn't import Django. Are you sure it's installed and "
            "available on your PYTHONPATH environment variable? Did you "
            "forget to activate a virtual environment?"
        ) from exc
    execute_from_command_line(sys.argv)


if __name__ == "__main__":
    main()

The code above will create telemetry wrappers inside django-outbox-pattern code and creates automatic spans with broker data.

The DjangoOutboxPatternInstrumentor can receive three optional parameters:

• trace_provider: The tracer provider to use in open-telemetry spans.
• publisher_hook: The callable function on publisher action to call before the original function call, use this to override, enrich the span or get span information in the main project.
• consumer_hook: The callable function on consumer action to call before the original function call, use this to override, enrich the span or get span information in the main project.

⚠️ The hook function will not raise an exception when an error occurs inside hook function, only a warning log is generated

Span Generated

save published {destination}

With the django-outbox-pattern, when you create on Published objects one object is saved and this saved is instrumentalized.

from django_outbox_pattern.models import Published

Published.objects.create(
    destination='destination',
    body={
        "random": "body"
    },
)
    

The outbox save span had save published {destination} name.

send {destination}

After save the object in Published model the publish command will get all pending messages and publish there to broker

python manage.py publish

The outbox publish span had publish {destination} name.

Consumer

Using the django-outbox-pattern, we create a simple consumer using subscribe management command, using this command we can see the consumer spans.

   python manage.py subscribe 'dotted.path.to.callback' 'destination' 'queue_name'

Consumer spans can generate up to three types:

• process {destination}
• ack {destination}
• nack {destination}

Supress django-outbox-pattern traces and metrics

When the flag OTEL_PYTHON_DJANGO_OUTBOX_PATTERN_INSTRUMENT has False value traces and metrics will not be generated. Use this to supress the django-outbox-pattern-instrumentation instrumentation.

HOW TO CONTRIBUTE ?

Look the contributing specs