Dockerfile with spark and jupyter instalation. It's the base for the engine image.
It contains:
- Jupyter 5.7.0
- Spark 2.3.2
- Toree 0.2.0-incubating-rc4
docker build -t srcd/jupyter-spark .
docker run -it -p 8888:8888 srcd/jupyter-spark
Jupyter listens in port 8888. You'll have to use the token provided in the log file to access the service.
In case you want to change the configuration to listen in another port or skip token check you can create a configuration file. For example:
RUN mkdir -p /root/.jupyter && \
echo "c.NotebookApp.token = ''" > ~/.jupyter/jupyter_notebook_config.py && \
echo "c.NotebookApp.open_browser = False" >> ~/.jupyter/jupyter_notebook_config.py && \
echo "c.NotebookApp.notebook_dir = '/home'" >> ~/.jupyter/jupyter_notebook_config.py && \
echo "c.NotebookApp.port = 8080" >> ~/.jupyter/jupyter_notebook_config.pyYou can modify Spark options at runtime by setting the SPARK_OPTS environment variable. For example:
docker run -it -p 8888:8888 -e 'SPARK_OPTS=--driver-memory=8g' srcd/jupyter-spark
Contributions are more than welcome, if you are interested please take a look to our Contributing Guidelines.
All activities under source{d} projects are governed by the source{d} code of conduct.
Apache License Version 2.0, see LICENSE.