使用 venv 运行 docker 容器时找不到 gunicorn

Question

I'm trying to run a Docker container which complains with the error message: /bin/sh: gunicorn: not found . Firing up the server locally without Docker works fine. And building the image also works fine. I'm new to docker so I don't know if anything looks weird in my Dockerfile..

My Dockerfile:

FROM python:3.7-alpine

RUN adduser -D teamreacher
WORKDIR /home/teamreacher

# copy and install dependencies
COPY ./requirements.txt requirements.txt
RUN python -m venv venv
RUN venv/bin/pip install --upgrade pip
RUN venv/bin/pip install -r requirements.txt

# copy the app
COPY . .
RUN chmod +x boot.sh

RUN chown -R teamreacher:teamreacher ./
USER teamreacher

# expose port and run server
EXPOSE 5000

RUN source venv/bin/activate
CMD gunicorn -b :5000 --access-logfile - --error-logfile - wsgi:app

And my requirements.txt:

Flask==1.0.2
Flask-RESTful==0.3.6
Flask-SQLAlchemy==2.3.2
Flask-JWT==0.3.2
Flask-Cors==3.0.7
gunicorn==19.9.0

Answer 1

A RUN command creates a layer, it's like running the command in a new shell. When it completes, the 'shell' exits. So any following commands will not be affected.

You can add a shell script (startup.sh) like,

#!/bin/sh
source venv/bin/activate
gunicorn -b :5000 --access-logfile - --error-logfile - wsgi:app

then CMD ["./startup.sh"]

PS:

There is little interest for using virtual env in docker container. A container is already an isolated environment and it's supposed to do only one thing.

Answer 2

My packages installation looked like this:

RUN pip install -r requirements.txt --target=/app/python

I had to remove the --target=/app/python then pod status started showing Running (got fixed).

So the fix for me was:

RUN pip install -r requirements.txt