Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

MyDocker@Paris-Saclay

MyDocker@Paris-Saclay is a virtual environment service that provides access to many scientific software (including Jupyter) from a simple web browser, notably for interactive computing, data processing, etc.

This service is open to students and staff of the University Paris‑Saclay. As an experimental offering, it is more generally open to institutions of the identity federation “Research and Higher Education” of Renater); contact us before any intensive use.

Access to environments

Your myDocker homepage gives you access to all the environments you have already used. The items below give you access to generic environments. If you are a teacher, you can create your own environments, or even propose that they be added here. Many other environments have been created by colleagues for all sorts of needs.

JupyterLab Interface

Clone of JupyterHub@Paris-Saclay (obsolete)

  • Software: Python and its standard libraries, SageMath, C++, R, ...

  • Interfaces: JupyterLab, Jupyter, Visual Studio Code or graphical desktop environment XFCE

  • Resources: global persistent personal folder, 2 CPUs, 4 GB RAM, shutdown after 20 minutes of inactivity

This multipurpose environment is exactly identical to the one deployed on the former JupyterHub@Paris-Saclay service and is intended to facilitate the transition.

Limitation: this environment has only been marginally updated since September 2023 and will remain as is. The versions of the installed software are therefore outdated. Moreover, it can be slow to load due to its size. Unless a specific need arises, we recommend using one of the more specialized environments.

🚀 Start

JupyterLab and Python 🚀 Start

This environment is designed for simple uses of Python, particularly for introductory programming and computing.

JupyterLab, Python and scientific libraries 🚀 Start

  • Interface: JupyterLab

  • Software: Python, NumPy, SciPy, MatPlotLib, Pandas, scikit-learn, conda, pip, ...

  • Resources: global persistent personal folder, 2 CPUs, 4 GiB RAM, shutdown after 20 minutes of inactivity

  • Environment configuration (restricted access)

  • Maintainer: Nicolas M. Thiéry

This environment is based on the image jupyter-scipy-notebook from the Jupyter community.

JupyterLab and PyTorch for deep learning 🚀 Start

  • Interface: JupyterLab

  • Software: Python, Numpy, SciPy, MatPlotLib, Pandas, scikit-learn, PyTorch, conda, pip, ...

  • Resources: global persistent personal folder, 4GB RAM, shutdown after 20 minutes of inactivity

  • Maintainer: Nicolas M. Thiéry

  • Environment configuration (restricted access)

This environment is based on the image jupyter/pytorch-notebook from the Jupyter community.

For technical reasons, there are currently no limits on CPU usage. Please do not abuse it!

JupyterLab and C++ 🚀 Start

  • Interface: JupyterLab, in French

  • Software: compilers (gcc, clang) and interpreter (cling) C++, Travo, ...

  • Resources: persistent global personal folder, 2 CPU, 4 GB RAM, shutdown after 20 minutes of inactivity, conversational agent

  • Courses:

    • «Introduction to Imperative Programming», L1 Math-Info, S2, Faculty of Sciences of Orsay

    • Modular Programming, L1 Math-Info, S2, Faculty of Sciences of Orsay

    • «Algorithms and Data Structures», L1 Math-Info, S2, Faculty of Sciences of Orsay

    • «Info 1», Polytech

    • ...

  • Maintainer: Nicolas M. Thiéry

  • Environment configuration (restricted access)

  • Image configuration

JupyterLab and SageMath 🚀 Start

  • Interface: JupyterLab, in French

  • Software: SageMath, Travo

  • Resources: persistent global personal folder, 2 CPU, 4 GB RAM, shutdown after 20 minutes of inactivity, conversational agent

  • Courses:

    • «Option C: Algebra and Symbolic Computation», Mathematics Aggregation, Faculty of Sciences of Orsay

    • «Combinatorics and Algebraic Computation», M1 MPRI, Faculty of Sciences of Orsay

    • «Math-Info Project», LDD1 Math-Info, S2, Faculty of Sciences of Orsay

    • ...

  • Maintainer: Nicolas M. Thiéry

  • Environment configuration (restricted access)

  • Image configuration

🚧 JupyterLab and Julia 🚀 Start

  • Interface: JupyterLab, in French

  • Software: Julia with packages SymPy, Plots, Statistics, DataFrames, DSP, Latexify

  • Resources: persistent personal folder and global personal folder (in shared/), 2 CPU, 4 GB RAM, shutdown after 30 minutes of inactivity

  • Course: ???

  • Maintainer: Nicolas M. Thiéry, Bastien Berret

  • Environment configuration (restricted access)

  • Image configuration

This environment is based on the image jupyter/pytorch-notebook from the Jupyter community.

🚧 JupyterLab and SQL 🚀 Start

  • Interface: JupyterLab

  • Software: PostgreSQL, Python and psycopg2 library

  • Resources: persistent personal folder with data directory in ~/pgsql/data, 2 CPU, 4 GB RAM

  • Services:

    • PostgreSQL (available at startup, psql --list shows the existing databases),

    • Web server (can be started with python -m http.server in the chosen folder; pages accessible via https://<mydocker-host>/proxy/8000)

  • Course: ???

  • Maintainer: Chiara Marmo

  • Environment configuration (restricted access)

This environment is based on the image jupyter/minimal-notebook.

🚧 JupyterLab and R 🚀 Start

This environment is based on the image jupyter/r-notebook.

This environment is under development. We contact for any suggestion of a classic R package to add is welcome; likewise, if you would like to use RStudio or Rcmd.

🚧 JupyterLab and LaTeX 🚀 Start

This environment is based on the image jupyter/minimal-notebook.

This environment is under development. Contact us for any suggestion of texlive packages to add.

This environment implements the collaborative features of JupyterLab and represents a first effort to develop a tool “like Overleaf”: the jupyter lab git extension is also integrated. It is an approximation of real-time collaboration which still has a number of limitations.

Interface: Visual Studio Code

VSCode and C++ 🚀 Start

  • Interface: Visual Studio Code

  • Software: gcc, clang, boost, ....

  • Resources: persistent personal folder 2GB, 2 CPU, 4GB RAM

  • Courses:

    • ...

  • Maintainer: Dominique Marcadet

  • Environment configuration (restricted access)

  • Image configuration

This environment is based on the image openvscode.

VSCode and Java 🚀 Start

  • Interface: Visual Studio Code

  • Software: openjdk-21, maven, gradle

  • Resources: persistent personal folder 2GB, 2 CPU, 4GB RAM

  • Courses:

    • ...

  • Maintainer: Dominique Marcadet

  • Environment configuration (restricted access)

  • Image configuration

This environment is based on the image openvscode.

VSCode and Rust 🚀 Start

This environment is based on the image openvscode.

On demand: VSCode and Javascript

Interface: graphical desktop with XFCE

XFCE, C++ and SFML 🚀 Start

This environment is based on the image WebTop image from LinuxServer.io.

XFCE, Eclipse and Java

  • Interface: Lightweight XFCE desktop environment

  • Software: Eclipse IDE for Java developers 2026-03, JDK 25, Maven 3.9.14

  • Maintainer: Dominique Marcadet

  • Image configuration

This environment is based on the image webtop

XFCE, DVWA and Burp 🚀 Start

This environment is based on the image webtop.

XFCE, Archi

This environment is based on the image webtop.

Interface: terminal

Terminal 🚀 Start

This environment is based on the alpine image

You can import / download files using the tsz and trz commands (does not work with all browsers, including firefox).

Terminal via SSH 🚀 Start

This environment is based on the alpine image

About the service

Essential features

  • Ubiquity: Use from a simple web browser.

  • Urbanization: authentication via the identity federation «Research and Higher Education» of Renater, or via integration as an external tool in Moodle (or other LMS compatible LTI).

  • Choice of environment, based on Jupyter or not
    With the possibility to define customized environments for finely selecting the software and resources used.

  • Scalability: sized for several hundred (thousands?) simultaneous users for light interactive uses (a few CPUs).
    For heavier needs, notably GPU for applications in AI, it is possible on a case‑by‑case basis to use the instance myDocker de Centrale-Supélec, whose servers are rented on demand. Contact us.

  • Data persistence.

  • Sovereign hosting.

  • AI: integration of a sovereign conversational agent (Aristote).

📜 History: from JupyterHub@Paris-Saclay to myDocker@Paris-Saclay

From 2017 to 2024, Paris-Saclay University provided its staff and students with a virtual environment service JupyterHub@Paris-Saclay giving access to a wide range of scientific software (including Jupyter). This service was hosted and operated by the Mesocenter DataCenter@UPSud of Paris-Saclay and was used by 8,000 members of higher education and research, including 5,000 at Paris-Saclay, with about a hundred daily users.

In parallel, CentraleSupélec had developed and deployed in 2019 a similar service, myDocker, with an emphasis on environment personalization, provision of heavier resources (GPU), scalability and control of the ecological footprint, for teaching artificial intelligence, software development, numerical simulation, modeling, etc.

Between November 2024 and January 2025, and with financial support from the CMA SaclAI-School, the two teams carried out a convergence between the two services, with an upgrade of myDocker to cover ongoing uses at Paris-Saclay and the deployment of myDocker@Paris-Saclay which takes over from JupyterHub@Paris-Saclay.

To facilitate the migration, a software environment strictly identical to that of JupyterHub@Paris-Saclay is available and the data have been transferred automatically. The migration was disrupted and delayed by a cyber‑attack, leading to two phases:

  • Late October 2024: data migration for users with an Adonis account @universite-paris-saclay.fr

  • First week of January 2025: data migration for the other users.

In cases where users had used both services during their coexistence, if a file existed on both services, the migration placed the most recent version on myDocker@Paris-Saclay.

myDocker@CentraleSupelec and myDocker@Paris-Saclay

The service myDocker@CentraleSupelec is maintained in parallel with myDocker@Paris-Saclay. It is hosted on sovereign servers rented on demand from OVH. It is intended primarily for students and staff of CentraleSupelec. As an experiment, and pending the definition of an appropriate economic model, access can be granted on a case-by-case basis via eCampus for heavy usage (including GPU access). Because of the rental model, it is necessary to schedule sessions.

🙏 Acknowledgments

The myDocker@Paris-Saclay service is co‑piloted by the Faculty of Science and Centrale Supélec, operated by CentraleSupélec, co‑developed by the DISI of CentraleSupélec and Illuin, co‑financed by the CMA SaclAI-School and hosted by the mesoscale centre DataCenter@UPSud.

📨🛠️ Contact and technical support

In case of a service outage, or for help organizing courses using myDocker, you can contact technical support: mydocker-upsaclay@listes.centralesupelec.fr.

⚠️ Known limitations

Authentication

Usage

Developments under study

🚧 Alternatives and similar projects

Some academic services:

Some commercial services:

MyDocker@Paris-Saclay (Planets)
Documentation