Details
-
Feature
-
Not Assigned
-
None
-
SRCnet
-
-
-
-
-
22.6
-
-
SRC-DM SRC-DataAccess SRC-SWdist SRCNet0.1 Team_CORAL data-lifecycle
Description
In the same way we needed to integrate as seamlessly as possible JHub and Rucio, we would like to do with other applications. A general case is when a user access to a container (which may have installed the tools required for the data analysis) to analyse a set of data stored in the data lake. We would like to avoid the manual steps required for the autenticantion (token capture) and the download of the dataset/data required by the user.
The idea is to provide the user with a container that would be responsible for authenticating the user as automatically as possible, download the data and make them available to the user. A way to do this is to launch the container using environmental variables: docker run .... -e user=mparra python test.py <datasetdid>
Then the container would take care of the above process.
This is related to https://jira.skatelescope.org/browse/SP-2654?src=confmacro and https://jira.skatelescope.org/browse/SP-2902
Who: Coral team with support from Cyan Team
What: Investigate how to automatize/facilitate the accesss to data in the data lake from containers (see link)
Why: Enable availability of access to Datalake from applications or services that are containerised.