Digital Repository Testbed Exploration

One touchstone of RLG’s mission is helping members meet the challenges of digital preservation. Much of that work has focused on defining the characteristics of a "trusted digital repository" where significant digital assets can be archived for the long term.

In 2002 RLG also got hands-on experience with digital archiving through a digital repository testbed project, conducted with a third-party provider of archiving services, JPMorgan’s i-VAULT! (SM).

Those early tests showed that i-VAULT! had promise as a provider of digital archiving services to the research community. RLG concluded at the time that using a third-party archiving service would be less economical than building an archive in-house—and that many RLG members would be likely to make similar calculations regarding archiving services.

• i-VAULT! home page

Background

[Adapted from [link to "/en/page.php?Page_ID=457"] RLG News issue 56, spring 2003]

"Digitization and digital life are growing exponentially," says RLG program officer Robin Dale. "Digital archiving is important now, but will be unavoidable for many institutions within two to three years."

RLG took important steps in 2002 to help its members meet the challenges of digital preservation. One of the ways it did so was through a digital repository testbed project, conducted with a third-party provider of archiving services, JPMorgan’s i-VAULT! (SM).

i-VAULT! has provided digital image archiving services for over 15 years. The company’s first clients were banks, who needed to store digital images of canceled checks, although the digital objects stored by customers have since expanded to include wills, deeds, titles, and other legal documents. JPMorgan i-VAULT! maintains two separate, redundant archive sites and is currently storing the equivalent of over 700 million document pages. (See www.jpmorgan.com/ivault for more information on this service from JPMorgan.)

After extensive discussions with i-VAULT!, RLG decided to test whether the company’s image archiving system would fit the bill as a third-party trusted digital archive for libraries, museums, and archives. "i-VAULT! was very attractive," says Dale, "because it could provide the archival storage function of OAIS," potentially saving RLG or its members the trouble and expense of building one of the significant components of an OAIS-compliant archive.

In the test project RLG took a ten-gigabyte subset of data from RLG Cultural Materials, stored these objects in a test repository provided by i-VAULT!, then retrieved data from the repository to verify that the objects had been successfully archived. In the process, says RLG software development manager Judith Bush, RLG learned a lot about the technical processes required to implement the OAIS model. For instance, RLG had to create submission information packages for transmission to i-VAULT!. The test team also needed to consider the access requirements and the Web interface that would be needed for future retrieval of those objects.

The tests showed that i-VAULT! has much promise as a digital repository. "It is a very robust service," says Linda West, RLG’s director of member programs and initiatives, "and could form the backbone of a certified digital repository. However, such services don’t come cheap." After considering "sunk costs" (those already spent on staff and servers), RLG concluded that using a third-party archiving service would be less economical than building an archive in-house. RLG expects to have over ten terabytes of data in RLG Cultural Materials alone within five years, notes RLG chief information officer Jack Grantham, and for that quantity of data, RLG judged that the cost of i-VAULT! would be too high to justify, since archival storage is only one component of a full-fledged digital repository.

Grantham believes that many institutions will make a similar calculation regarding third-party archiving services, especially when considering labor costs. Universities, for instance, have access to existing IT staff resources—and inexpensive undergraduate labor pools. Because these costs are low (or aren’t included in budgets for in-house archiving projects), third-party archiving services look costly by comparison. "One huge issue that any certified digital archive will have to face when approaching big universities is that you can’t create a pricing model that will be convincing to them," says Grantham.

However, the initial steps taken by RLG and i-VAULT! bode well for the future. "It’s encouraging that there’s a large, world-renowned vendor out there that’s willing to work with our community to create a digital archiving service that might work for us," says Dale. Eventually, the price of third-party solutions such as i-VAULT! may converge with the amount that libraries and other institutions are willing to pay for such services. "At some point in the future people are going to have to bite the bullet and make the investment in archiving," Dale says. "i-VAULT! might be a reasonable solution for institutions who cannot build and maintain their own archival storage."