When dealing with data produced or used by HPC applications, you should keep in mind two separated separate issues: where to put them and how to move them. This means "Data Storage" and "Data transfer".
Where: The management of a large amount of data (BigData) is not a trivial task: you can use a normal FileSystem (FS) based area or you should consider to describe your data with with metadata, in order to better organize them and share among your colleagues (REPO).
How: The transfer of big data sets from or to an HPC cluster with classical tools like scp is not straightforward because various constraints could limit bandwidth and execution time.
...