Accessing Datasets

Searching for datasets in DAS

In order to find a given CMS dataset in DBS, one can use either the DAS web-interface or the stand-alone dasgoclient

Requesting a dataset transfer

Datasets needed for use in analyses that will be run at Purdue can be requested through Rucio.

• Policy
• Managing datasets via Rucio

Access information inside a dataset

• Commands to Operate on EOS Data

Speed-up repeated dataset access via XCache

One can significantly improve the speed of dataset reading during repeated access of the same files (e.g. during developing or testing analysis code) by temporarily caching those files locally at the site in high-performant NVMe disks. This way subsequent reads of a datafile - be it from the site's own storage system, or remote from another site - will be much faster.

• How to use (User's perspective)
• How it works (Technical Details)