Hey everyone! I’m also excited to (finally!) share my technical blog post detailing how my repo above is able to hot-load code changes in a container and how poetry can be used for package management. Any feedback is welcome! • Blog Post • Github repo

This is great @Alex Service, wish i had this a month ago 🙂 . I've had success with a single stage dockerfile using basically the same folder setup. the key i found is to copy the files after running the install step, which will use the caching mechanism.

FROM python:3.8-slim-buster

ENV POETRY_VERSION=1.1.13

# system dependencies
RUN pip install "poetry==$POETRY_VERSION"

# copy requirements to cache in docker layer
RUN mkdir /opt/code
WORKDIR /opt/code

# Project initialization
COPY pyproject.toml poetry.lock ./

# Install poetry to ensure Python dependencies are correct
RUN poetry config virtualenvs.create false \
    && poetry install --no-dev

COPY . .

I use the dagster cloud ci/cd and mount the repo as a package, and it works great

Glad you like it! And that’s an interesting approach; does hot-loading of code still work, or does the container still need restarted on changes? I’m guessing it still needs restarted, but is quicker than it would otherwise be

yeah, it still needs to be restarted, but i don't really do any interactive development using the container -- just use the poetry venv and `poetry run dagit --package-name my_package. `` The container is really just for deployment in my case

Hi @Alex Service it’s a great blog post! I’ve recently created this Github Discussion to help share projects that are using Dagster. Feel free to share your post and repo there!

Done! Thanks @yuhan 🙂