Towards sustainable, enterprise-scale Research Data Storage and Management C. Gardiner 1 , S. Gray 2 , S. Price 2 , D. Boyd 3 , V. Knight 2 , D. Steer 2 and B. Cregan 1 1 Advanced Computing Research Centre, 2 IT Services R&D/ILRT and 3 Library University of Bristol, 8-10 Berkeley Square, Bristol BS8 1HH, UK 1 Introduction The University of Bristol’s road map towards sustainable research data management builds on the success of its petascale Research Data Storage Facility (RDSF) through the data.bris 1 programme of projects to establish a University-wide repository service. The eventual data.bris Research Data Repository Service is designed to meet the growing demand from researchers for data-related support and will better align the University with the requirements of funders and publishers by fulfilling the following functions. • Construct, maintain and track the usage of an externally accessible digital repository consisting of the University’s research data. Datasets will be cross-referenced with related research outputs (e.g. journal articles) to be listed in Pure, the institutional Research Information System (RIS). • Deliver and maintain an on-demand, professional training course, open to all research staff. This course will be focused on helping staff meet the Research Data Management (RDM) mandates which have recently been introduced by many research funding bodies. • Applicants to RCUK funding councils are expected to provide a detailed Data Management Plan (DMP) explaining how data management issues will be addressed. The repository service will support and guide research staff during the pre-award, DMP authoring phase and offer ongoing support to help address RDM requirements during and at the close of research activities. In this paper we share what we consider to be the critical success factors of the RDSF and describe how a similar model is being applied in data.bris with the aim of achieving effective research data management across the enterprise. We conclude by reporting on some of the open questions around the challenging problem of sustainable research data curation. 2 Research Data Storage The University set up a central High Performance Computing (HPC) facility in 2006 and it quickly became apparent that there was significant demand for a means of storing the data generated by re- searchers using the HPC facility, in particular the climate modellers and particle physicists. As we started to quantify demand, we also discovered that researchers from other disciplines had major data storage requirements, particularly in the Arts and Medical faculties, where large images and/or videos were being produced. Data was being stored on a range of devices, often without reliable backup and with no long term preservation strategy for the data. An initial requirement was identified for a resilient, secure, long term facility, providing one petabyte of storage initially, but with the ability to be scaled up as demand grows. The University recognised the need for such a facility and investment of £2 million was agreed in 2007 to establish the Research Data Storage Facility (RDSF). It was decided the facility should be solely for research data, so that we could focus on the specific issues which arise when storing research data; other kinds of data such as administrative data, student email, teaching materials have their own issues. 1 data.bris – http://data.bris.ac.uk 1