Communications - April 2012

fice of Science and Technology Policy released a Request for Information, soliciting comments regarding public access to digital data resulting from federally funded research in November 2011.

Also, some of the research funded under NSF’s Sustainable Digital Data Preservation and Access Network Partners (DataNet) program is beginning to bear significant fruit. Although Spengler says the DataNet projects were and are intended to be exemplars and fairly restrictive prototypes due to limited funding, her NSF colleague Rob Pennington says DataNet award-ees are working with other researchers eager to find ways to share data across domains and disciplines.

One standout example of this is iRODS (integrated Rule-Orient-ed Data System), developed by the Data Intensive Cyber Environments (DICE) research group at the University of North Carolina (UNC) and the University of California, San Diego. Institutions spanning disciplines from climatology to social sciences are adopting the innovative data grid tool. The NSF awarded iRODS developers $8 million in September 2011 to build a policy-driven national data management infrastructure, motivated by the discrete data management requirements of the NSF’s Ocean Observatories Initiative, NSF’s Consortium of Universities for Advancement of Hydrologic Science, engineering projects in education, CAD/CAM/ CAE archives, the genomic databases of the iPlant collaborative, the H.W. Odum Institute for Research in Social Science at UNC, and NSF’s Science of Learning Centers.

iRODS has also been adopted by scientific data centers worldwide, including astronomical observatories in Canada and France, climate centers in the U.S., and at the Sanger Institute genomics databases in the U.K. It is also in use in the U.S. National Archives and Records Administration’s Transcontinental Persistent Archives Prototype.

iRODS is the successor to the pioneering Storage Resource Broker (SRB) architecture. DICE Director Reagan Moore says iRODS’s rule engine-based architecture makes the distributed management of a data grid much sim-

there is little to
no national-level
coordination of
data preservation
standards, even
in the u.K.’s
well-documented
research council
guidelines.

pler than the hard-coded SRB architecture, can serve as the sort of reporting tool that demonstrates researchers are meeting their mandated data management plan outlines—and is also the sort of policy-based system that melds the concepts of data management and data preservation for whatever duration is required.

“We make the assertion that any
data-management application really
consists of the policies you’re apply-
ing in order to validate assertions
about what you’ve done,” Moore says.
“So if I build a preservation environ-
ment, my policies are related to au-
thenticity, integrity, chain of custody,
and original arrangement.”
Moore says the rule-based iRODS
architecture makes it possible for us-
ers to tailor which policies apply to a
given action without having to rewrite
any code, whereas server-side com-
mands in SRB were hard coded. These
rules are applied in a platform-agnos-
tic manner through any number of 254
microservices selected as pertinent by
any user community.

At least one project has already successfully replicated a recognized sample archive, the Harvard IQSS-developed Dataverse, using iRODS. Researchers at UNC’s Odum Institute performed a Dataverse-to-iRODS transfer using the Open Archives Initiative Protocol for Metadata Harvesting and the compatible Data Documentation Initiative standard, plus XML.

“The result is an accurate copy of a
Dataverse archive inside iRODS,” ac-
cording to the UNC authors, “which
data grid administrators can preserve
over the long term by, for example,
replicating the information to many
geographically distributed storage re-
sources.”
Phil Butcher, head of information
technology at the Sanger Institute, says
the organization’s iRODS installation
has run smoothly. He believes fund-
ing agencies should make themselves
aware of the details of such ground-
breaking technologies, even if com-
prehensive national and international
management and preservation policies
are not possible.

Further Reading

Beagrie, N., Lavoie, B., and Woollard, M. Keeping Research Data Safe 2, JISC, Bristol, U.K., April 2010.. Chiang, G., Clapham, P., Qi, G., Sale, K., and Coates, G. Implementing a genomic data management system using iRODS in the Wellcome Trust Sanger Institute, BMC Bioinformatics 12, 361, Sept. 9, 2011.

Neuroth, H., Strathmann, S., and Vlaeminck, S. Digital preservation needs of scientific communities: The example of Göttingen University, Proceedings of the 12th European Conference on Research and Advanced Technology for Digital Libraries, Aarhus, Denmark, Sept. 14–19, 2008. Rothenberg, J. and Hoorens, S. Enabling Long-term Access to Scientific, Technical and Medical Data Collections, RAnD Europe, Cambridge, U.K., 2010. Ward, J.H., de Torcy, A., Chua, M., and Crabtree, J. Extracting and ingesting DDI metadata and digital objects from a data archive into the iRODS extension of the nARA TPAP using the OAI-PMh, 5th IEEE International Conference on e-Science, Oxford, UK, Dec. 9–11, 2009.