'Docker image python-alpine fails when using lib
I have the following Dockerfile:
FROM python:3.7-alpine
WORKDIR /msa-app/
EXPOSE 3000
ENV LD_LIBRARY_PATH="/usr/local/lib"
ENV LD_RUN_PATH="/usr/local/lib"
ENV SAPNWRFC_HOME="/usr/local/sap/nwrfcsdk"
ENV PATH="${PATH}:/usr/sap/nwrfcsdk/lib"
COPY nwrfcsdk/nwrfcsdk /usr/local/sap/nwrfcsdk
COPY nwrfcsdk/nwrfcsdk.conf /etc/ld.so.conf.d/nwrfcsdk.conf
COPY msa-app /msa-app/
RUN chmod +x /msa-app/entrypoint.sh
ENTRYPOINT /msa-app/entrypoint.sh
With the following entrypoint.sh (I put everything in entrypoint due to proxy issues inside the network that I'm using - don't blame me, it's the only way that I found to fix an issue):
#!/bin/sh
#Creates and activates virtual env
echo "*** Activating Python virtual environment"
python -m venv /msa-app/venv && source /msa-app/venv/bin/activate
echo "*** Installing pip requirements"
pip install --upgrade pip && pip install -r /msa-app/requirements.txt
echo "*** Installing pyrfc"
pip install /msa-app/pyrfc-1.9.95-cp37-cp37m-linux_x86_64.whl
echo "*** Configuring SAPNWRFC PyRFC"
ln -s /usr/bin/python3 python
mkdir -p /etc/ld.so.conf.d/
mkdir -p /usr/sap/
ldconfig /usr/local/lib
#Starts gunicorn for flask
echo "*** Starting application"
cd /msa-app && gunicorn -w 2 -b 0.0.0.0:3000 run:app
But when I try to run the container I get the following error:
ImportError: Error loading shared library libsapnwrfc.so: No such file or directory (needed by /msa-app/venv/lib/python3.7/site-packages/pyrfc/_pyrfc.cpython-37m-x86_64-linux-gnu.so)
But, this error DOESN'T happens when I use the normal non-alpine Python Docker Image. What am I doing wrong? Thank you.
Solution 1:[1]
This configuration worked for me when faced with this error.
# For more information, please refer to https://aka.ms/vscode-docker-python
FROM python:3-alpine3.10
RUN apk add build-base
RUN pip install --upgrade pip
# Copy the requirements.txt to the root of the docker image
COPY ./requirements.txt .
# Keeps Python from generating .pyc files in the container
ENV PYTHONDONTWRITEBYTECODE=1
# Turns off buffering for easier container logging
ENV PYTHONUNBUFFERED=1
ENV TZ=$TIME_ZONE
ENV DJANGO_PROJECT_NAME core
ENV EXPOSED_CONTAINER_PORT 2092
# Add header files required to install dependencies in the requirements.txt file
RUN apk add --update --no-cache --virtual .tmp gcc libc-dev linux-headers python3-dev musl-dev unixodbc unixodbc-dev \
&& apk add --no-cache mariadb-dev mariadb-client postgresql-libs postgresql-dev \
&& apk add jpeg-dev zlib-dev libjpeg
# Install pip requirements
RUN pip install -r requirements.txt
# Install Pillow for image management
RUN pip install Pillow
# Install MS-SQL dependencies
RUN pip install mssql-django
# Install postgres engine adapter
RUN pip install psycopg2
# Install pytest for running unit-tests
RUN pip install pytest
# Delete the temporarily directory used for downloading the required image header.
# This is done to reduce on the image size and also to remove these redundant files
RUN apk del .tmp
# This is where static files will be stored in the container. -p makes is used to create all directories that don't exist
RUN mkdir -p /app/static
# Change directory and make app as the current working directory
WORKDIR /app
COPY . /app
CMD find .
RUN pytest /app/tests/
# Creates a non-root user with an explicit UID and adds permission to access the /app folder
# For more info, please refer to https://aka.ms/vscode-docker-python-configure-containers
RUN adduser -u 5678 --disabled-password --gecos "" appuser && chown -R appuser /app
USER appuser
# Expose the container's point in order to be accessed outside the container
EXPOSE $EXPOSED_CONTAINER_PORT:$EXPOSED_CONTAINER_PORT
# Copy the entry point script to the container root directory
COPY ./entrypoint.sh /
ENTRYPOINT ["sh", "/entrypoint.sh" ]
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
Solution | Source |
---|---|
Solution 1 | Umar Kayondo |