CODEWITHSUNDEEP
pythondjangodockerdocker-composepip
mirodil
mirodil

Reputation: 537

Installing a package in docker

I am using docker in django project, and installed packages which are in req.txt.
During project I needed to install a package and did it using docker-compose exec web pip install 'package' and docker-compose up -d --build, it installed in docker but I cannot use it in my project that not installed in project.
Question:

Dockerfile:

FROM python:3.8

ENV PYTHONDONTWRITEBYTECODE 1
ENV PYTHONUNBUFFERED 1

WORKDIR /code

COPY requirements.txt /code/
RUN pip install -r requirements.txt

COPY . /code/

docker-compose.yml:

version: '3.9'

services:
  web:
    build: .
    command: python /code/manage.py runserver 0.0.0.0:8000

    volumes:
      - .:/code

    ports:
      - 8000:8000
    depends_on:
      - db

  db:
    image: postgres:11
    volumes:
      - postgres_data:/var/lib/postgresql/data/
    environment:
      - "POSTGRES_HOST_AUTH_METHOD=trust"

volumes:
  postgres_data:

Upvotes: 0

Views: 2731

Answers (1)

David Maze
David Maze

Reputation: 158647

Here's the standard workflow I'd use here.

First of all, make sure you have a normal working non-Docker development environment.

python3 -m venv vpy
. vpy/bin/activate
pip install -e .

Then edit your setup.cfg file to include the new package and update the virtual environment.

$EDITOR setup.cfg
# add install_requires= package
pip install -e .

Make sure your application works properly.

pytest
run_the_app  # calling a setuptools console_scripts= script

Update the package lock file.

pip freeze > requirements.txt

All of this so far is still on your host, outside of Docker. Now you can rebuild the image to run integration tests:

docker-compose build
docker-compose up -d
curl http://localhost:8000

Note that this setup is the totally normal Python dependency workflow, plus a final "build your Docker image" at the end. There is nothing Docker-specific in it, and you do not "install a dependency in Docker" or "update the host source code from the deployed image". You could do a similar thing with newer tools like Pipenv instead of plain virtual environments.

(You do not need a Compose volumes: block to hide the code in the image, nor should you need a Compose command: line to override the Dockerfile's CMD.)

Upvotes: 1

Related Questions