Details
-
Feature
-
Won't have (this time)
-
None
-
None
-
SRCnet
-
-
-
17.6
Description
Context: The datalake-as-a-service (DLaaS) is a prototype which integrates storage that forms part of a Rucio datalake prototype with JupyterHub, providing a user interface to schedule the staging of data within a user's Jupyter Notebook environment. It was initially developed at CERN within the ESCAPE project, but has since been replicated as part of the SRC prototyping activities (https://confluence.skatelescope.org/display/SRCSC/SKAO+Data+Lake+as+a+Service+Deployment+-+Setup+Guide).
Summary: Add the ability to upload new data to the datalake from the Rucio-JupyterLab extension.
Currently, the UI elements for interacting with Rucio from JupyterHub are in the Rucio jupyterlab-extension. When this was first developed (~1 year ago) the token-based upload functionality was lacking in Rucio, so a complex workaround using non-deterministic RSEs was developed. This is no longer necessary as token-based uploads now work, so it should be possible to simplify the jupyterlab-extension to use the existing Rucio upload method.
AC: UI element in Rucio jupyterlab-extension can be used to upload data into the data lake, at an RSE of the user's choosing.