14 · Survey Results: Executive Summary
The most used access storage solution is a network
file system (43 responses or 72%). External media li-
brary and local/attached storage each received 27
responses (45%). One respondent noted that they use
Amazon Cloud and hosted Open Repository. Another
uses a local DSpace instance, the California Digital
Library’s web archiving service, and a university
system-wide open access repository. Other solutions
include the use of a local implementation of a Fedora
repository, YouSendIt online file sharing software in
combination with e-mail, and shared IT servers.
The most common back up storage solution is a
network file system (44 responses or 76%), followed by
external media library (31 or 53%), local/attached stor-
age (23 or 40%), and distributed systems (16 or 28%).
Other solutions include a combination of Amazon
Cloud and hosted Open Repository, the California
Digital Library’s Merritt Repository, redundant stor-
age managed by campus and library IT, and physical
tape storage.
Network file systems are used most for dark stor-
age (26 responses or 52%), with distributed comput-
ing/storage systems second (19 or 38%). External me-
dia library and local/attached storage were not far
behind at 16 and 14 responses, respectively. Other
dark storage solutions include the California Digital
Library’s Merritt Repository, the Chronopolis digital
preservation network, the Isilon commercial storage
platform, redundant storage managed by campus
and library IT, and virtual and physical tape storage.
One respondent stated that rather than dark storage,
their institution uses Fedora as an asset management
system and copies files to “replicated storage for long-
term preservation, with appropriate preservation
metadata and restricted access.”
Estimating Storage Needs and Costs
Twenty-six of the responding libraries (59%) estimate
future digital storage needs and costs based on past
and current usage and/or planned growth. Three
noted that storage is allocated on a case-by-case basis.
Some respondents have yet to implement methods of
estimating storage needs and costs. Others are in the
process of developing such methods.
Respondents described a variety of approaches to
estimating storage needs and costs. One is conducting
a longitudinal analysis of trends in digital storage
growth. Another will scale future digital storage
needs to the “development of campus department
operations.” Another currently uses costs of disks,
storage devices, and backups as the basis for total
cost estimates and is looking at moving to endow-
ment-based storage cost models in the future. One
respondent anticipates using the L.I.F.E. (Life Cycle
Information for E-Literature) model developed by
University College London (UCL) and the British
Library for estimating curation costs, including the
cost of storage.
One institution estimates space needs based on
“past collecting volume +a 20% inflator +any known
collections we anticipate receiving.” Another esti-
mates required storage needs based on average file
size for a particular type of record and then estimates
costs based on the current market value of storage,
“usually at the TB level.”
The most detailed response described the institu-
tion’s attempt to estimate storage needs by tracking
historical usage and growth, contrasting those with
earlier projections, and categorizing data by type to
identify growth areas. Thus far, the respondent ob-
serves that “consumption generally increases by a
factor of 2 to 4 within a 12–18 month period,” but any
projection can change when unexpected projects or
changes in the organization occur.
Access and Discovery
The survey asked which delivery methods the li-
brary uses to provide access to born-digital materials.
Two-thirds of respondents provide online access to
a digital repository system. Just under half provide
in-library access on a dedicated workstation. Users
who bring their PCs to 22 of the responding libraries
can access born-digital materials stored on portable
media. Eighteen respondents (28%) use third-party
systems such as CONTENTdm, Archive-It, Dropbox,
and YouTube to share materials with researchers.
There is not one, single repository system being
used either to manage or provide access to born-
digital materials. Most respondents use open source
repository software for both management and access
functions. Twenty-eight institutions report using se-
cure file system storage to manage collections but only
Previous Page Next Page