django_app_gunicorn_statsd_prometheus

Example Django application

See src for the application code and top level README for the description of this repo from a functionality point of view.

Building Docker image

The Python 3 based Dockerfile uses an Alpine Linux base image and expects the application source code to be volume mounted at /application when run:

FROM python:3.7-alpine
ADD . /application
WORKDIR /application
RUN set -e; \
	apk add --no-cache --virtual .build-deps \
		gcc \
		libc-dev \
		linux-headers \
	; \
	pip install -r src/requirements.txt; \
	apk del .build-deps;
EXPOSE 8000
VOLUME /application

CMD gunicorn --bind 0.0.0.0:8000  demo.wsgi

The last statement shows how we are running the application via gunicorn with 5 worker processes.

To build the image:

$ docker build -t amitsaha/django_app_1 -f Dockerfile.py3 .

Running the application

We can just run the web application as follows:

$ docker run  -ti -p 8000:8000 -v `pwd`/src:/application amitsaha/django_app_1

Bringing up the web application, along with prometheus

The docker-compse.yml brings up the webapp service which is our web application using the image amitsaha/flask_app_1 we built above. The docker-compose-infra.yml file brings up the statsd service which is the statsd exporter, prometheus service and also starts the grafana service which is available on port 3000. The config directory contains a prometheus.yml file which sets up the targets for prometheus to scrape. The scrape configuration looks as follows:

# A scrape configuration containing exactly one endpoint to scrape:
# Here it's Prometheus itself.
scrape_configs:
  # The job name is added as a label `job=<job_name>` to any timeseries scraped from this config.
  - job_name: 'prometheus'

    # Override the global default and scrape targets from this job every 5 seconds.
    scrape_interval: 5s

    # metrics_path defaults to '/metrics'
    # scheme defaults to 'http'.

    static_configs:
         - targets: ['localhost:9090']
  - job_name: 'webapp'

    # Override the global default and scrape targets from this job every 5 seconds.
    scrape_interval: 5s

    # metrics_path defaults to '/metrics'
    # scheme defaults to 'http'.
    static_configs:
        - targets: ['statsd:9102']

Prometheus scrapes itself, which is the first target above. The second target is the statsd exporter on port 9102.

Since these services are running via docker-compose, statsd automatically resolves to the IP of the statsd exporter container.

To bring up all the services:

$ docker-compose -f docker-compose.yml -f docker-compose-infra.yml up