Wavefront Django SDK

This SDK provides support for reporting out of the box metric, histograms and tracing from your Django based application. That data is reported to Wavefront via proxy or direct ingestion. That data will help you understand how your application is performing in production.

Install

pip install wavefront_django_sdk_python

Usage

Configure settings.py of your application to install Django SDK as follows:

# setting.py

from wavefront_pyformance.wavefront_reporter import WavefrontDirectReporter, WavefrontProxyReporter
from wavefront_sdk.common import ApplicationTags
from wavefront_django_sdk import DjangoTracing
from wavefront_opentracing_sdk import reporting, WavefrontTracer

INSTALLED_APPS = [
   '...',
   'wavefront_django_sdk',
   '...'
]

MIDDLEWARE = [
   'wavefront_django_sdk.middleware.WavefrontMiddleware',
   '...'
]

SOURCE = "{SOURCE}"

APPLICATION_TAGS = ApplicationTags(
   application="{APP_NAME}",
   service="{SERVICE_NAME}",
   cluster="{CLUSTER_NAME}",  # Optional
   shard="{SHARD_NAME}," , # Optional
   custom_tags={"location": "Oregon", "env": "Staging"}  # Optional
)

# Sending data via Direct Ingestion
WF_REPORTER = WavefrontDirectReporter(
   server="{ADDRESS}",
   token="{TOKEN}",
   reporting_interval=5,  # Optional, default value is 10 secs
   source=SOURCE,
   tags={"application": APPLICATION_TAGS.application}
).report_minute_distribution()

# Or, Sending data via Proxy
WF_REPORTER = WavefrontProxyReporter(
   host="{HOST}",
   port=2878,  # Optional, Wavefront Proxy running on 2878 by default
   reporting_interval=5,  # Optional, default value is 10 secs
   source=SOURCE,
   tags={"application": APPLICATION_TAGS.application}
).report_minute_distribution()

span_reporter = reporting.WavefrontSpanReporter(
   client=WF_REPORTER.wavefront_client,
   source=SOURCE,
)

OPENTRACING_TRACE_ALL = True  # Optional, default value is False

OPENTRACING_TRACING = DjangoTracing(WavefrontTracer(
   reporter=span_reporter, application_tags=APPLICATION_TAGS))

Out of the box metrics and histograms for your Django based application.

Assume you have the following API in your Django Application:

# urls.py
from django.urls import path
from . import views
 
urlpatterns = [
    path('style/<slug:id>/make', views.make_shirts, name="style/{id}/make")
]
 
# view.py
from django.http import HttpResponse
 
def make_shirts(request, id):
    return HttpResponse("completed", status=200)

Request Gauges

Granular Response related metrics

Granular Response related histograms

Overall Response related metrics

This includes all the completed requests that returned a response (i.e. success + errors).

Overall Error Response related metrics

This includes all the completed requests that resulted in an error response (that is HTTP status code of 4xx or 5xx).

Tracing Spans

Every span will have the operation name as span name, start time in millisec along with duration in millisec. The following table includes all the rest attributes of generated tracing spans.