Curation in the Cloud White Paper
Download the white paper (PDF)
In recent years the hype around cloud computing as a route to delivering a dazzling array of services has (to quote Gartner) been 'deafening'. But much of the focus has been on using cloud infrastructure to create, process and analyse data, often in relation to business information and in the context of either the virtualisation and green ICT agendas, or for the purpose of providing flexible computational capacity for actively undertaking research work. There has been less opportunity to consider the potential of this technology to address the data curation and long-term preservation needs of researchers and research organisations. Especially in light of emerging UK government-led directives on ensuring long-term open access to publicly funded research, it is more important than ever that organisations have a clear sense of all the choices that are available to them when investing in this critical area.
JISC in association with the Kindura Project (King's College London) organised a 2-day workshop to assess the potential and practicalities of using cloud-based solutions for the curation and long-term preservation of digital materials, focusing particularly on data that originates from research or that supports research processes. At the workshop participants discussed a draft white paper prepared by the Digital Curation Centre. A revised version of this white paper is now publicly available - please see below. (It is also available from the Digital Curation Centre website).
Digital curation involves a wide range of activities, many of which may be suitable for deployment within a cloud environment. These range from infrequent, resource-intensive tasks which will benefit from the ability to rapidly provision resources, to day-to-day collaborative activities which can be facilitated by networked cloud services. Associated benefits are offset by risks such as loss of data or service level, legal and governance incompatibilities and transfer bottlenecks. There is considerable variability across both risks and benefits according to the service and deployment models being adopted and the context in which activities are performed. Some risks, such as legal liabilities, are mitigated by the use of alternatives, for example, private cloud models, but this is typically at the expense of benefits such as resource elasticity and economies of scale. The Infrastructure as a Service (IaaS) model may provide a basis on which more specialised software services may be provided.
There is considerable work to be done in helping institutions understand the cloud and its associated costs, risks and benefits, and how these compare to their current working methods, in order that the most beneficial uses of cloud technologies may be identified. Specific proposals, echoing recent work coordinated by EPSRC and JISC are the development of advisory, costing and brokering services to facilitate appropriate cloud deployments, the exploration of opportunities for certifying or accrediting cloud preservation providers, and the targeted publicity of outputs from pilot studies to the full range of stakeholders within the curation lifecycle, including data creators and owners, repositories, institutional IT support professionals and senior managers.