hcstubbe / dev_einvironment

A repository for development containers for RStudio and JupyterLab

Docker development environment

Important notes:

• The data within each container and the database container does not persist, when the stack is removed! For development save changes using a version control system and export data and results!
• Alternatively, you can mount persistent folders of the hosts computers file system into the containers by modifying the YAML-files like this.

Django testing

• To test on mariadb (DO NOT USE THE PRODUCTION DATABASE FOR TESTING) run the following on the database server:
  • mariadb --user=root --password=danjgocou # start cli
  • GRANT ALL PRIVILEGES ON test_mydbdevdanjgo.* TO 'userdanjgo'; # change permissions

JupyterLab and RStudio

The following runs the JupyterLab and RStudio containers including a MariaDB SQL database server (without GPU enabled) for TensorFlow and general R pipelines

• Install the Docker engine or Docker Desktop (be aware of the license requirements!)
• Start with: docker-compose -f dev-stack.yml up -d
• Then, open on the local host
  • RStudio: http://localhost:9090/ (user: rstudio, password: test123; change in dev-stack.yml!)
  • JupyterLab: http://localhost:9090/ (you can find the token by running docker exec -it [CONTAINER ID] jupyter server list on the jupyter-lab container.
• Shutdown with: docker-compose -f dev-stack.yml down

JupyterLab and RStudio with GPU acceleration

The following runs the JupyterLab and RStudio containers including a MariaDB SQL database server with GPU enabled for TensorFlow and general R pipelines

• Setup for GPU support using NVIDIA CUDA compatible images and GPUs:
  • Installation guide for Linux/Docker-engine with CUDA support
  • Installation on Windows 10/11 with WSL and Docker Desktop:
    • Install WSL2
    • Install Docker Desktop (be aware of the license requirements!)
    • Install Nvidia drivers for Windows
• Start with GPU support docker-compose -f dev-stack-gpu.yml up -d (this requires an NVIDIA GPU and the respective drivers/installation)
• Then, open on the local host
  • RStudio: http://localhost:9090/ (user: rstudio, password: test123; change in dev-stack.yml!)
  • JupyterLab: http://localhost:9090/ (you can find the token by running docker exec -it [CONTAINER ID] jupyter server list on the jupyter-lab container.
• Shutdown with: docker-compose -f dev-stack-gpu.yml down

The django stack

The following runs the JupyerLab and RStudio containers for developing Django web apps including a MariaDB server.

• Install the Docker engine or Docker Desktop (be aware of the license requirements!)
• Start with: docker-compose -f dj-tf-notebook/dev-stack.yml up -d
• Then, open on the local host to run
  • RStudio: http://localhost:9090/ (user: rstudio, password: test123; change in dev-stack.yml!)
  • JupyterLab: http://localhost:9090/ (you can find the token by running docker exec -it [CONTAINER ID] jupyter server list on the jupyter-lab container.
  • Django development server: run python manage.py runserver 0.0.0.0:8000 in the JupyerLab terminal and open http://localhost:9090/
• Shutdown with: docker-compose -f dj-tf-notebook/dev-stack.yml down