7
0
mirror of https://github.com/ChristianLight/tutor.git synced 2024-05-31 13:20:48 +00:00
tutor/tutor/templates/build/openedx/Dockerfile

268 lines
11 KiB
Docker
Raw Normal View History

###### Minimal image with base system requirements for most stages
v11.0.0 (2020-12-09) - 💥[Improvement] Upgrade Open edX to Koa - 💥 Setting changes: - The ``ACTIVATE_HTTPS`` setting was renamed to ``ENABLE_HTTPS``. - Other ``ACTIVATE_*`` variables were all renamed to ``RUN_*``. - The ``WEB_PROXY`` setting was removed and ``RUN_CADDY`` was added. - The ``NGINX_HTTPS_PORT`` setting is deprecated. - Architectural changes: - Use Caddy as a web proxy for automated SSL/TLS certificate generation: - Nginx no longer listens to port 443 for https traffic - The Caddy configuration file comes with a new ``caddyfile`` patch for much simpler SSL/TLS management. - Configuration files for web proxies are no longer provided. - Kubernetes deployment no longer requires setting up a custom Ingress resource or custom manager. - Gunicorn and Whitenoise are replaced by uwsgi: this increases boostrap performance and makes it no longer necessary to mount media folders in the Nginx container. - Replace memcached and rabbitmq by redis. - Additional features: - Make it possible to disable all plugins at once with ``plugins disable all``. - Add ``tutor k8s wait`` command to wait for a pod to become ready - Faster, more reliable static assets with local memory caching - Deprecation: proxy files for Apache and Nginx are no longer provided out of the box. - Removed plugin `{{ patch (...) }}` statements: - "https-create", "k8s-ingress-rules", "k8s-ingress-tls-hosts": these are no longer necessary. Instead, declare your app in the "caddyfile" patch. - "local-docker-compose-nginx-volumes": this patch was primarily used to serve media assets. The recommended is now to serve assets with uwsgi.
2020-09-17 10:53:14 +00:00
FROM docker.io/ubuntu:20.04 as minimal
LABEL maintainer="Overhang.io <contact@overhang.io>"
2017-07-03 10:39:19 +00:00
v11.0.0 (2020-12-09) - 💥[Improvement] Upgrade Open edX to Koa - 💥 Setting changes: - The ``ACTIVATE_HTTPS`` setting was renamed to ``ENABLE_HTTPS``. - Other ``ACTIVATE_*`` variables were all renamed to ``RUN_*``. - The ``WEB_PROXY`` setting was removed and ``RUN_CADDY`` was added. - The ``NGINX_HTTPS_PORT`` setting is deprecated. - Architectural changes: - Use Caddy as a web proxy for automated SSL/TLS certificate generation: - Nginx no longer listens to port 443 for https traffic - The Caddy configuration file comes with a new ``caddyfile`` patch for much simpler SSL/TLS management. - Configuration files for web proxies are no longer provided. - Kubernetes deployment no longer requires setting up a custom Ingress resource or custom manager. - Gunicorn and Whitenoise are replaced by uwsgi: this increases boostrap performance and makes it no longer necessary to mount media folders in the Nginx container. - Replace memcached and rabbitmq by redis. - Additional features: - Make it possible to disable all plugins at once with ``plugins disable all``. - Add ``tutor k8s wait`` command to wait for a pod to become ready - Faster, more reliable static assets with local memory caching - Deprecation: proxy files for Apache and Nginx are no longer provided out of the box. - Removed plugin `{{ patch (...) }}` statements: - "https-create", "k8s-ingress-rules", "k8s-ingress-tls-hosts": these are no longer necessary. Instead, declare your app in the "caddyfile" patch. - "local-docker-compose-nginx-volumes": this patch was primarily used to serve media assets. The recommended is now to serve assets with uwsgi.
2020-09-17 10:53:14 +00:00
ENV DEBIAN_FRONTEND=noninteractive
2017-12-26 00:16:35 +00:00
RUN apt update && \
apt install -y build-essential curl git language-pack-en
ENV LC_ALL en_US.UTF-8
{{ patch("openedx-dockerfile-minimal") }}
###### Install python with pyenv in /opt/pyenv and create virtualenv in /openedx/venv
FROM minimal as python
# https://github.com/pyenv/pyenv/wiki/Common-build-problems#prerequisites
2020-10-16 21:22:02 +00:00
RUN apt update && \
apt install -y libssl-dev zlib1g-dev libbz2-dev \
libreadline-dev libsqlite3-dev wget curl llvm libncurses5-dev libncursesw5-dev \
xz-utils tk-dev libffi-dev liblzma-dev python-openssl git
feat: upgrade to Maple - A shared cookie domain between lms and cms is no longer recommended: https://github.com/edx/edx-platform/blob/master/docs/guides/studio_oauth.rst - refactor: clean mounted data folder in lms/cms. In Lilac, the bind-mounted lms/data and cms/data folders are a mess because new folders are created there for every new course organisation. These folders are empty. As far as we know they are useless... With this change we move these folders to a dedicated "modulestore" subdirectory; which corresponds better to the initial intent of the fs_root setting. - fix: frontend failure during login to the lms. See: https://github.com/openedx/build-test-release-wg/issues/104 - feat: move all forum-related code to a dedicated plugin. Forum is an optional feature, and as such it deserves its own plugin. Starting from Maple, users will be able to install the forum from https://github.com/overhangio/tutor-forum/ - migrate from DCS_* session cookie settings to SESSION_*. That's because edx-platform no longer depends on django-cookies-samesite. Close https://github.com/openedx/build-test-release-wg/issues/110 - get rid of tons of deprecation warnings in the lms/cms - feat: make it possible to point to themed assets. Cherry-picking this change makes it possible to point to themed assets with a theme-agnostic url, notably from MFEs. - Install all official plugins as part of the `tutor[full]` package. - Don't print error messages about loading plugins during autocompletion. - Prompt for image building when upgrading from one release to the next. - Add `tutor local start --skip-build` option to skip building Docker images. Close #450. Close #545.
2021-10-18 09:43:40 +00:00
ARG PYTHON_VERSION=3.8.12
ENV PYENV_ROOT /opt/pyenv
feat: upgrade to Maple - A shared cookie domain between lms and cms is no longer recommended: https://github.com/edx/edx-platform/blob/master/docs/guides/studio_oauth.rst - refactor: clean mounted data folder in lms/cms. In Lilac, the bind-mounted lms/data and cms/data folders are a mess because new folders are created there for every new course organisation. These folders are empty. As far as we know they are useless... With this change we move these folders to a dedicated "modulestore" subdirectory; which corresponds better to the initial intent of the fs_root setting. - fix: frontend failure during login to the lms. See: https://github.com/openedx/build-test-release-wg/issues/104 - feat: move all forum-related code to a dedicated plugin. Forum is an optional feature, and as such it deserves its own plugin. Starting from Maple, users will be able to install the forum from https://github.com/overhangio/tutor-forum/ - migrate from DCS_* session cookie settings to SESSION_*. That's because edx-platform no longer depends on django-cookies-samesite. Close https://github.com/openedx/build-test-release-wg/issues/110 - get rid of tons of deprecation warnings in the lms/cms - feat: make it possible to point to themed assets. Cherry-picking this change makes it possible to point to themed assets with a theme-agnostic url, notably from MFEs. - Install all official plugins as part of the `tutor[full]` package. - Don't print error messages about loading plugins during autocompletion. - Prompt for image building when upgrading from one release to the next. - Add `tutor local start --skip-build` option to skip building Docker images. Close #450. Close #545.
2021-10-18 09:43:40 +00:00
RUN git clone https://github.com/pyenv/pyenv $PYENV_ROOT --branch v2.2.2 --depth 1
RUN $PYENV_ROOT/bin/pyenv install $PYTHON_VERSION
v11.0.0 (2020-12-09) - 💥[Improvement] Upgrade Open edX to Koa - 💥 Setting changes: - The ``ACTIVATE_HTTPS`` setting was renamed to ``ENABLE_HTTPS``. - Other ``ACTIVATE_*`` variables were all renamed to ``RUN_*``. - The ``WEB_PROXY`` setting was removed and ``RUN_CADDY`` was added. - The ``NGINX_HTTPS_PORT`` setting is deprecated. - Architectural changes: - Use Caddy as a web proxy for automated SSL/TLS certificate generation: - Nginx no longer listens to port 443 for https traffic - The Caddy configuration file comes with a new ``caddyfile`` patch for much simpler SSL/TLS management. - Configuration files for web proxies are no longer provided. - Kubernetes deployment no longer requires setting up a custom Ingress resource or custom manager. - Gunicorn and Whitenoise are replaced by uwsgi: this increases boostrap performance and makes it no longer necessary to mount media folders in the Nginx container. - Replace memcached and rabbitmq by redis. - Additional features: - Make it possible to disable all plugins at once with ``plugins disable all``. - Add ``tutor k8s wait`` command to wait for a pod to become ready - Faster, more reliable static assets with local memory caching - Deprecation: proxy files for Apache and Nginx are no longer provided out of the box. - Removed plugin `{{ patch (...) }}` statements: - "https-create", "k8s-ingress-rules", "k8s-ingress-tls-hosts": these are no longer necessary. Instead, declare your app in the "caddyfile" patch. - "local-docker-compose-nginx-volumes": this patch was primarily used to serve media assets. The recommended is now to serve assets with uwsgi.
2020-09-17 10:53:14 +00:00
RUN $PYENV_ROOT/versions/$PYTHON_VERSION/bin/python -m venv /openedx/venv
2017-07-03 10:39:19 +00:00
###### Install Dockerize to wait for mysql DB availability
FROM minimal as dockerize
# https://github.com/powerman/dockerize/releases
ARG DOCKERIZE_VERSION=v0.16.0
RUN dockerize_url="https://github.com/powerman/dockerize/releases/download/$DOCKERIZE_VERSION/dockerize-linux-$(uname -m | sed 's@aarch@arm@')" \
&& echo "Downloading dockerize from $dockerize_url" \
&& curl --fail --location --output /usr/local/bin/dockerize $dockerize_url \
&& chmod a+x /usr/local/bin/dockerize
###### Checkout edx-platform code
FROM minimal as code
ARG EDX_PLATFORM_REPOSITORY={{ EDX_PLATFORM_REPOSITORY }}
ARG EDX_PLATFORM_VERSION={{ EDX_PLATFORM_VERSION }}
RUN mkdir -p /openedx/edx-platform && \
git clone $EDX_PLATFORM_REPOSITORY --branch $EDX_PLATFORM_VERSION --depth 1 /openedx/edx-platform
WORKDIR /openedx/edx-platform
2017-07-03 10:39:19 +00:00
# Identify tutor user to apply patches using git
RUN git config --global user.email "tutor@overhang.io" \
&& git config --global user.name "Tutor"
2022-04-20 17:05:06 +00:00
{%- if patch("openedx-dockerfile-git-patches-default") %}
# Custom edx-platform patches
{{ patch("openedx-dockerfile-git-patches-default") }}
2022-04-20 17:05:06 +00:00
{%- else %}
# Patch edx-platform
# Fix broken "Pages" view in Studio
# https://github.com/openedx/edx-platform/pull/30550
RUN curl -fsSL https://github.com/open-craft/edx-platform/commit/3d54f284f82b61e693ad652d8d6e46a226fcb36d.patch | git am
# Fix xblock ajax handler vulnerability
# https://github.com/overhangio/edx-platform/tree/overhangio/sec-xblock-ajax
RUN curl -fsSL https://github.com/overhangio/edx-platform/commit/3f0f9eed42.patch | git am
2022-04-20 17:05:06 +00:00
{%- endif %}
{# Example: RUN curl -fsSL https://github.com/openedx/edx-platform/commit/<GITSHA1> | git am #}
{{ patch("openedx-dockerfile-post-git-checkout") }}
###### Download extra locales to /openedx/locale/contrib/locale
FROM minimal as locales
ARG OPENEDX_I18N_VERSION={{ OPENEDX_COMMON_VERSION }}
RUN cd /tmp \
&& curl -L -o openedx-i18n.tar.gz https://github.com/openedx/openedx-i18n/archive/$OPENEDX_I18N_VERSION.tar.gz \
&& tar xzf /tmp/openedx-i18n.tar.gz \
&& mkdir -p /openedx/locale/contrib \
&& mv openedx-i18n-*/edx-platform/locale /openedx/locale/contrib \
&& rm -rf openedx-i18n*
###### Install python requirements in virtualenv
FROM python as python-requirements
ENV PATH /openedx/venv/bin:${PATH}
ENV VIRTUAL_ENV /openedx/venv/
RUN apt update && apt install -y software-properties-common libmysqlclient-dev libxmlsec1-dev libgeos-dev
# Note that this means that we need to reinstall all requirements whenever there is a
# change in edx-platform, which sucks. But there is no obvious alternative, as we need
# to install some packages from edx-platform.
COPY --from=code /openedx/edx-platform /openedx/edx-platform
WORKDIR /openedx/edx-platform
# Install the right version of pip/setuptools
# https://pypi.org/project/setuptools/
# https://pypi.org/project/pip/
# https://pypi.org/project/wheel/
RUN pip install setuptools==62.1.0 pip==22.0.4 wheel==0.37.1
# Install base requirements
RUN pip install -r ./requirements/edx/base.txt
2017-07-03 10:39:19 +00:00
v11.0.0 (2020-12-09) - 💥[Improvement] Upgrade Open edX to Koa - 💥 Setting changes: - The ``ACTIVATE_HTTPS`` setting was renamed to ``ENABLE_HTTPS``. - Other ``ACTIVATE_*`` variables were all renamed to ``RUN_*``. - The ``WEB_PROXY`` setting was removed and ``RUN_CADDY`` was added. - The ``NGINX_HTTPS_PORT`` setting is deprecated. - Architectural changes: - Use Caddy as a web proxy for automated SSL/TLS certificate generation: - Nginx no longer listens to port 443 for https traffic - The Caddy configuration file comes with a new ``caddyfile`` patch for much simpler SSL/TLS management. - Configuration files for web proxies are no longer provided. - Kubernetes deployment no longer requires setting up a custom Ingress resource or custom manager. - Gunicorn and Whitenoise are replaced by uwsgi: this increases boostrap performance and makes it no longer necessary to mount media folders in the Nginx container. - Replace memcached and rabbitmq by redis. - Additional features: - Make it possible to disable all plugins at once with ``plugins disable all``. - Add ``tutor k8s wait`` command to wait for a pod to become ready - Faster, more reliable static assets with local memory caching - Deprecation: proxy files for Apache and Nginx are no longer provided out of the box. - Removed plugin `{{ patch (...) }}` statements: - "https-create", "k8s-ingress-rules", "k8s-ingress-tls-hosts": these are no longer necessary. Instead, declare your app in the "caddyfile" patch. - "local-docker-compose-nginx-volumes": this patch was primarily used to serve media assets. The recommended is now to serve assets with uwsgi.
2020-09-17 10:53:14 +00:00
# Install django-redis for using redis as a django cache
# https://pypi.org/project/django-redis/
RUN pip install django-redis==5.2.0
v11.0.0 (2020-12-09) - 💥[Improvement] Upgrade Open edX to Koa - 💥 Setting changes: - The ``ACTIVATE_HTTPS`` setting was renamed to ``ENABLE_HTTPS``. - Other ``ACTIVATE_*`` variables were all renamed to ``RUN_*``. - The ``WEB_PROXY`` setting was removed and ``RUN_CADDY`` was added. - The ``NGINX_HTTPS_PORT`` setting is deprecated. - Architectural changes: - Use Caddy as a web proxy for automated SSL/TLS certificate generation: - Nginx no longer listens to port 443 for https traffic - The Caddy configuration file comes with a new ``caddyfile`` patch for much simpler SSL/TLS management. - Configuration files for web proxies are no longer provided. - Kubernetes deployment no longer requires setting up a custom Ingress resource or custom manager. - Gunicorn and Whitenoise are replaced by uwsgi: this increases boostrap performance and makes it no longer necessary to mount media folders in the Nginx container. - Replace memcached and rabbitmq by redis. - Additional features: - Make it possible to disable all plugins at once with ``plugins disable all``. - Add ``tutor k8s wait`` command to wait for a pod to become ready - Faster, more reliable static assets with local memory caching - Deprecation: proxy files for Apache and Nginx are no longer provided out of the box. - Removed plugin `{{ patch (...) }}` statements: - "https-create", "k8s-ingress-rules", "k8s-ingress-tls-hosts": these are no longer necessary. Instead, declare your app in the "caddyfile" patch. - "local-docker-compose-nginx-volumes": this patch was primarily used to serve media assets. The recommended is now to serve assets with uwsgi.
2020-09-17 10:53:14 +00:00
# Install uwsgi
# https://pypi.org/project/uWSGI/
feat: upgrade to Maple - A shared cookie domain between lms and cms is no longer recommended: https://github.com/edx/edx-platform/blob/master/docs/guides/studio_oauth.rst - refactor: clean mounted data folder in lms/cms. In Lilac, the bind-mounted lms/data and cms/data folders are a mess because new folders are created there for every new course organisation. These folders are empty. As far as we know they are useless... With this change we move these folders to a dedicated "modulestore" subdirectory; which corresponds better to the initial intent of the fs_root setting. - fix: frontend failure during login to the lms. See: https://github.com/openedx/build-test-release-wg/issues/104 - feat: move all forum-related code to a dedicated plugin. Forum is an optional feature, and as such it deserves its own plugin. Starting from Maple, users will be able to install the forum from https://github.com/overhangio/tutor-forum/ - migrate from DCS_* session cookie settings to SESSION_*. That's because edx-platform no longer depends on django-cookies-samesite. Close https://github.com/openedx/build-test-release-wg/issues/110 - get rid of tons of deprecation warnings in the lms/cms - feat: make it possible to point to themed assets. Cherry-picking this change makes it possible to point to themed assets with a theme-agnostic url, notably from MFEs. - Install all official plugins as part of the `tutor[full]` package. - Don't print error messages about loading plugins during autocompletion. - Prompt for image building when upgrading from one release to the next. - Add `tutor local start --skip-build` option to skip building Docker images. Close #450. Close #545.
2021-10-18 09:43:40 +00:00
RUN pip install uwsgi==2.0.20
{{ patch("openedx-dockerfile-post-python-requirements") }}
# Install private requirements: this is useful for installing custom xblocks.
COPY ./requirements/ /openedx/requirements
RUN cd /openedx/requirements/ \
&& touch ./private.txt \
&& pip install -r ./private.txt
{% for extra_requirements in OPENEDX_EXTRA_PIP_REQUIREMENTS %}RUN pip install '{{ extra_requirements }}'
{% endfor %}
###### Install nodejs with nodeenv in /openedx/nodeenv
FROM python as nodejs-requirements
ENV PATH /openedx/nodeenv/bin:/openedx/venv/bin:${PATH}
# Install nodeenv with the version provided by edx-platform
RUN pip install nodeenv==1.6.0
RUN nodeenv /openedx/nodeenv --node=12.13.0 --prebuilt
2018-12-25 23:08:06 +00:00
# Install nodejs requirements
ARG NPM_REGISTRY={{ NPM_REGISTRY }}
COPY --from=code /openedx/edx-platform/package.json /openedx/edx-platform/package.json
COPY --from=code /openedx/edx-platform/package-lock.json /openedx/edx-platform/package-lock.json
WORKDIR /openedx/edx-platform
RUN npm clean-install --verbose --registry=$NPM_REGISTRY
2017-07-03 10:39:19 +00:00
###### Production image with system and python requirements
FROM minimal as production
# Install system requirements
2020-10-16 21:22:02 +00:00
RUN apt update && \
v11.0.0 (2020-12-09) - 💥[Improvement] Upgrade Open edX to Koa - 💥 Setting changes: - The ``ACTIVATE_HTTPS`` setting was renamed to ``ENABLE_HTTPS``. - Other ``ACTIVATE_*`` variables were all renamed to ``RUN_*``. - The ``WEB_PROXY`` setting was removed and ``RUN_CADDY`` was added. - The ``NGINX_HTTPS_PORT`` setting is deprecated. - Architectural changes: - Use Caddy as a web proxy for automated SSL/TLS certificate generation: - Nginx no longer listens to port 443 for https traffic - The Caddy configuration file comes with a new ``caddyfile`` patch for much simpler SSL/TLS management. - Configuration files for web proxies are no longer provided. - Kubernetes deployment no longer requires setting up a custom Ingress resource or custom manager. - Gunicorn and Whitenoise are replaced by uwsgi: this increases boostrap performance and makes it no longer necessary to mount media folders in the Nginx container. - Replace memcached and rabbitmq by redis. - Additional features: - Make it possible to disable all plugins at once with ``plugins disable all``. - Add ``tutor k8s wait`` command to wait for a pod to become ready - Faster, more reliable static assets with local memory caching - Deprecation: proxy files for Apache and Nginx are no longer provided out of the box. - Removed plugin `{{ patch (...) }}` statements: - "https-create", "k8s-ingress-rules", "k8s-ingress-tls-hosts": these are no longer necessary. Instead, declare your app in the "caddyfile" patch. - "local-docker-compose-nginx-volumes": this patch was primarily used to serve media assets. The recommended is now to serve assets with uwsgi.
2020-09-17 10:53:14 +00:00
apt install -y gettext gfortran graphviz graphviz-dev libffi-dev libfreetype6-dev libgeos-dev libjpeg8-dev liblapack-dev libmysqlclient-dev libpng-dev libsqlite3-dev libxmlsec1-dev lynx ntp pkg-config rdfind && \
2020-10-16 21:22:02 +00:00
rm -rf /var/lib/apt/lists/*
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
# From then on, run as unprivileged "app" user
# Note that this must always be different from root (APP_USER_ID=0)
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
ARG APP_USER_ID=1000
RUN if [ "$APP_USER_ID" = 0 ]; then echo "app user may not be root" && false; fi
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
RUN useradd --home-dir /openedx --create-home --shell /bin/bash --uid ${APP_USER_ID} app
USER ${APP_USER_ID}
COPY --from=dockerize /usr/local/bin/dockerize /usr/local/bin/dockerize
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
COPY --chown=app:app --from=code /openedx/edx-platform /openedx/edx-platform
COPY --chown=app:app --from=locales /openedx/locale /openedx/locale
COPY --chown=app:app --from=python /opt/pyenv /opt/pyenv
COPY --chown=app:app --from=python-requirements /openedx/venv /openedx/venv
COPY --chown=app:app --from=python-requirements /openedx/requirements /openedx/requirements
COPY --chown=app:app --from=nodejs-requirements /openedx/nodeenv /openedx/nodeenv
COPY --chown=app:app --from=nodejs-requirements /openedx/edx-platform/node_modules /openedx/edx-platform/node_modules
ENV PATH /openedx/venv/bin:./node_modules/.bin:/openedx/nodeenv/bin:${PATH}
ENV VIRTUAL_ENV /openedx/venv/
WORKDIR /openedx/edx-platform
# Re-install local requirements, otherwise egg-info folders are missing
RUN pip install -r requirements/edx/local.in
# Create folder that will store lms/cms.env.yml files, as well as
# the tutor-specific settings files.
RUN mkdir -p /openedx/config ./lms/envs/tutor ./cms/envs/tutor
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
COPY --chown=app:app revisions.yml /openedx/config/
ENV LMS_CFG /openedx/config/lms.env.yml
ENV CMS_CFG /openedx/config/cms.env.yml
ENV REVISION_CFG /openedx/config/revisions.yml
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
COPY --chown=app:app settings/lms/*.py ./lms/envs/tutor/
COPY --chown=app:app settings/cms/*.py ./cms/envs/tutor/
# Copy user-specific locales to /openedx/locale/user/locale and compile them
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
RUN mkdir /openedx/locale/user
COPY --chown=app:app ./locale/ /openedx/locale/user/locale/
RUN cd /openedx/locale/user && \
django-admin compilemessages -v1
# Compile i18n strings: in some cases, js locales are not properly compiled out of the box
# and we need to do a pass ourselves. Also, we need to compile the djangojs.js files for
# the downloaded locales.
RUN ./manage.py lms --settings=tutor.i18n compilejsi18n
RUN ./manage.py cms --settings=tutor.i18n compilejsi18n
# Copy scripts
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
COPY --chown=app:app ./bin /openedx/bin
RUN chmod a+x /openedx/bin/*
ENV PATH /openedx/bin:${PATH}
{{ patch("openedx-dockerfile-pre-assets") }}
# Collect production assets. By default, only assets from the default theme
# will be processed. This makes the docker image lighter and faster to build.
# Only the custom themes added to /openedx/themes will be compiled.
# Here, we don't run "paver update_assets" which is slow, compiles all themes
# and requires a complex settings file. Instead, we decompose the commands
# and run each one individually to collect the production static assets to
# /openedx/staticfiles.
ENV NO_PYTHON_UNINSTALL 1
ENV NO_PREREQ_INSTALL 1
# We need to rely on a separate openedx-assets command to accelerate asset processing.
# For instance, we don't want to run all steps of asset collection every time the theme
# is modified.
RUN openedx-assets xmodule \
&& openedx-assets npm \
&& openedx-assets webpack --env=prod \
&& openedx-assets common
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
COPY --chown=app:app ./themes/ /openedx/themes/
RUN openedx-assets themes \
&& openedx-assets collect --settings=tutor.assets \
# De-duplicate static assets with symlinks
&& rdfind -makesymlinks true -followsymlinks true /openedx/staticfiles/
2017-07-03 10:39:19 +00:00
# Create a data directory, which might be used (or not)
RUN mkdir /openedx/data
2017-12-26 00:16:35 +00:00
# service variant is "lms" or "cms"
ENV SERVICE_VARIANT lms
ENV DJANGO_SETTINGS_MODULE lms.envs.tutor.production
2017-12-26 00:16:35 +00:00
{{ patch("openedx-dockerfile") }}
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
EXPOSE 8000
###### Intermediate image with dev/test dependencies
FROM production as development
# Install useful system requirements (as root)
USER root
RUN apt update && \
apt install -y vim iputils-ping dnsutils telnet \
&& rm -rf /var/lib/apt/lists/*
USER app
# Install dev python requirements
RUN pip install -r requirements/edx/development.txt
RUN pip install ipdb==0.13.4 ipython==7.27.0
# Add ipdb as default PYTHONBREAKPOINT
ENV PYTHONBREAKPOINT=ipdb.set_trace
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
# Recompile static assets: in development mode all static assets are stored in edx-platform,
# and the location of these files is stored in webpack-stats.json. If we don't recompile
# static assets, then production assets will be served instead.
RUN rm -r /openedx/staticfiles && \
mkdir /openedx/staticfiles && \
openedx-assets webpack --env=dev
{{ patch("openedx-dev-dockerfile-post-python-requirements") }}
# Default django settings
ENV DJANGO_SETTINGS_MODULE lms.envs.tutor.development
feat: run all services as unprivileged containers With this change, containers are no longer run as "root" but as unprivileged users. This is necessary in some environments, notably some Kubernetes clusters. To make this possible, we need to manually fix bind-mounted volumes in docker-compose. This is pretty much equivalent to the behaviour in Kubernetes, where permissions are fixed at runtime if the volume owner is incorrect. Thus, we have a consistent behaviour between docker-compose and Kubernetes. We achieve this by bind-mounting some repos inside "*-permissions" services. These services run as root user on docker-compose and will fix the required permissions, as per build/permissions/setowner.sh These services simply do not run on Kubernetes, where we don't rely on bind-mounted volumes. There, we make use of Kubernete's built-in volume ownership feature. With this change, we get rid of the "openedx-dev" Docker image, in the sense that it no longer has its own Dockerfile. Instead, the dev image is now simply a different target in the multi-layer openedx Docker image. This makes it much faster to build the openedx-dev image. Because we declare the APP_USER_ID in the dev/docker-compose.yml file, we need to pass the user ID from the host there. The only way to achieve that is with a tutor config variable. The downside of this approach is that the dev/docker-compose.yml file is no longer portable from one machine to the next. We consider that this is not such a big issue, as it affects the development environment only. We take this opportunity to replace the base image of the "forum" image. There is now no need to re-install ruby inside the image. The total image size is only decreased by 10%, but re-building the image is faster. In order to run the smtp service as non-root, we switch from namshi/smtp to devture/exim-relay. This change should be backward-compatible. Note that the nginx container remains privileged. We could switch to nginxinc/nginx-unprivileged, but it's probably not worth the effort, as we are considering to get rid of the nginx container altogether. Close #323.
2021-09-23 10:04:19 +00:00
CMD ./manage.py $SERVICE_VARIANT runserver 0.0.0.0:8000
###### Final image with production cmd
FROM production as final
2017-07-03 10:39:19 +00:00
# Run server
v11.0.0 (2020-12-09) - 💥[Improvement] Upgrade Open edX to Koa - 💥 Setting changes: - The ``ACTIVATE_HTTPS`` setting was renamed to ``ENABLE_HTTPS``. - Other ``ACTIVATE_*`` variables were all renamed to ``RUN_*``. - The ``WEB_PROXY`` setting was removed and ``RUN_CADDY`` was added. - The ``NGINX_HTTPS_PORT`` setting is deprecated. - Architectural changes: - Use Caddy as a web proxy for automated SSL/TLS certificate generation: - Nginx no longer listens to port 443 for https traffic - The Caddy configuration file comes with a new ``caddyfile`` patch for much simpler SSL/TLS management. - Configuration files for web proxies are no longer provided. - Kubernetes deployment no longer requires setting up a custom Ingress resource or custom manager. - Gunicorn and Whitenoise are replaced by uwsgi: this increases boostrap performance and makes it no longer necessary to mount media folders in the Nginx container. - Replace memcached and rabbitmq by redis. - Additional features: - Make it possible to disable all plugins at once with ``plugins disable all``. - Add ``tutor k8s wait`` command to wait for a pod to become ready - Faster, more reliable static assets with local memory caching - Deprecation: proxy files for Apache and Nginx are no longer provided out of the box. - Removed plugin `{{ patch (...) }}` statements: - "https-create", "k8s-ingress-rules", "k8s-ingress-tls-hosts": these are no longer necessary. Instead, declare your app in the "caddyfile" patch. - "local-docker-compose-nginx-volumes": this patch was primarily used to serve media assets. The recommended is now to serve assets with uwsgi.
2020-09-17 10:53:14 +00:00
CMD uwsgi \
--static-map /static=/openedx/staticfiles/ \
--static-map /media=/openedx/media/ \
--http 0.0.0.0:8000 \
--thunder-lock \
--single-interpreter \
--enable-threads \
--processes=${UWSGI_WORKERS:-2} \
--buffer-size=8192 \
--wsgi-file $SERVICE_VARIANT/wsgi.py
{{ patch("openedx-dockerfile-final") }}