Data Commons and Jupyter integration in CERNBox

About

CERNBox is CERN's open-source Enterprise File Sync&Share service and uses the Open Cloud Mesh (OCM) protocol to support federated sharing across independently operated sites.

EOSC Data Commons will provide building blocks that connect discovery, selection, and execution of research-data workflows, including services such as the EOSC Matchmaker and the EOSC Data Player. Through the project, such services are expected to enable the researcher to select a dataset (or set of datasets) and deliver it to a working environment for further processing: in this context, CERNBox plays the role of the working environment for researchers, leveraging its storage such as CephFS and its integration with JupyterLab.

In this talk, we present the integration work being done in CERNBox, where discovery results are turned into a package for processing and delivered into a user-accessible workspace via OCM-based federation mechanisms. The objective is to make it straightforward for researchers to bring selected datasets into CERNBox, run analysis in Jupyter-based environments and HPC resources close to the data, and then publish resulting outputs to repositories with minimal manual data handling.

00:00:00 Slide 1

00:00:11 Slide 2

00:00:27 Slide 3

00:01:06 Slide 4

00:03:22 Slide 5

00:05:19 Slide 6

00:05:56 Slide 7

00:10:45 Slide 8