Challenges for Online Research Data Enclaves

Thursday, October 28, 2021 | 12:15PM–1:00PM ET
Viewing Location: Online
Session Type: Poster Session
Delivery Format: Poster
“Virtual” data enclaves for analyzing restricted research data are growing in popularity because the sensitive data stay on the server and access can suspended at any time. The enables researchers to work remotely while maintaining security. Remote servers have been particularly important during the COVID-19 pandemic of 2020–21. Moreover, the enclaves act as collaboratories where projects share space for programs, documents, and data. Despite these advantages, enclaves still face important challenges. Two areas are HPC and identity management. (1) Incorporate high-performance computing (HPC) into the security enclave. Not all projects require HPC, so the challenge is to provide HPC as needed and at a scalable price. Any HPC must maintain security requirements including preventing end users from copying files from the server. (2) Some research data have restrictions on simultaneous access. In some circumstances, researchers must not be able to access data from multiple sources at the same time. The restriction prevents data from being merged or combined. The challenge is that most system operate under authentication and authorization. A system must provide access to data from source 1 or data from source 2, but not data from sources 1 and 2 (XOR). Different login credentials is contrary to best practices for identity management. Setting up different secure systems for each data source is not practical or cost effective.

Presenters

  • John Marcotte

    Archivist, University of Michigan-Ann Arbor