Scheduled maintenance of HPC notebook platform#

The notebook platform at will be offline for maintenance on July 4th at 10:00 (CEST). This operation will last 30 minutes and all connections to existing lab sessions will be interrupted.

Updated on 04/07/2023

10:30 The maintenance of is complete. All systems were upgraded successfully and the service is back online.

Once the maintenance is complete, users of the notebook platform will have to log in again through the web interface at to start a new lab session. Anybody with an open lab at the time of the maintenance will find their notebooks in the same state as the time of the disconnect. However we cannot guarantee data safety for data related to processes running in your notebooks. Therefore, we recommend all users to manually close their labs before this maintenance starts.

This maintenance will improve the usability of the panel to select the resources and environment of your lab sessions. We gathered user feedback in the past months and have reorganized and enhanced the available JupyterLab environments:

  • The dask environment has been expanded into a DataScience environment where core data science packages are pre-loaded (e.g. numpy, scipy, pandas, …) and that also allows to display interactive graphs with matplotlib directly in your notebooks.

  • The nglview environment has been expanded into a Molecules environment by adding a lab extension to display molecules directly in your notebooks with 3Dmol and also including all features of the new DataScience environment.

  • The version of JupyterLab launched by all environments will remain unchanged. Hence, impact of this update to user workflows in the lab will be minimal.

The table of available resources at the bottom of the selection panel has been redesigned. The information about total CPU cores and GPUs in the HPC has been replaced with information about available slots for 1 CPU core, 2 CPU cores, 4 CPU cores and 1 GPU jobs. This format will more accurately show the available resources to run lab sessions in the cluster.

Finally, this maintenance operation will also increase performance of the system running the notebook platform and update many components of its underlying operating system. JupyterHub, the main component of the notebook platform interface, will be upgraded as well.

If you have any questions or comments about this maintenance on the notebook platform, please contact VUB-HPC Support