Cloud Storage

DuraCloud Offers Redundant Storage Across Multiple Cloud Providers

To assist preservation and archiving efforts, a new cloud offering allows academics to store data with multiple cloud providers via a single interface.

As universities wrestle with the implications and opportunities of the cloud, it's clear that many academics have reservations about entrusting valuable data--in many cases, a life's worth of work--to a distant storage provider. Indeed, many academics distrust any system over which they don't have personal control--where they can't see and touch the actual hardware.

Such an approach is not always feasible in today's computing and economic environment, however, nor is it necessarily the safest way to protect data. A new service, known as DuraCloud, aims to assuage academic fears about the cloud by offering a more robust choice: providing redundant storage across multiple cloud providers.

DuraCloud is the latest product from DuraSpace, a nonprofit best known for sustaining Fedora and Dspace, two of the most dominant open source software platforms for long-term management of digital assets. Clients that sign on with DuraCloud can opt to store their data--as well as redundant copies--with one or more prominent cloud providers including Amazon, RackSpace, and the San Diego Supercomputer Center (SDSC) at the University of California, San Diego.

Clients can add or drop providers at any time, and may move content across them seamlessly. And regardless of how many providers they choose, universities have just one relationship to manage: the one with DuraCloud.

“We have a very long-term goal and mission of working with institutions--academic, government and cultural heritage--to be able to steward content for the long run," explains Michele Kimpton, CEO of DuraSpace. "DuraCloud was developed with this mission in mind. We are well positioned to mitigate risk because we’re transparent, we’re not in it for profit, and we’re able to look at it through the eyes of the community. We have very different motivations.”

Called a digital-preservation pioneer by the Library of Congress, Kimpton touts the system's transparency and security. “There are many cloud providers, but none of them are offering the services that we are," she notes. "There’s a whole list of benefits to using DuraCloud, all without having to wrangle service details or manage technical operations.”

Features offered by DuraCloud:

  • It's designed to support replication and backup activities, preservation and archiving, repository backup and multimedia access.
  • Clients use DuraCloud software as a service (SaaS) to manage settings and content.
  • From a customized dashboard, clients can manage content, set up sub-accounts, and control permissions.
  • Users can move copies of content into the cloud and store them with different providers.
  • All content is automatically checked for health, and corrupt files are replaced.
  • Storage terms and the location of data are completely transparent.
  • Migration strategies can be executed on a client’s behalf.
  • Clients work exclusively with DuraCloud for one fixed rate.
  • Stored data can be shared and streamed.

It's a feature set that certainly appealed to the University of Michigan's Institute for Social Research, home to the world’s largest social science data archive: the Inter-University Consortium for Political and Social Research (ICPSR). “It’s just good archival practice to store data in multiple locations,” says Bryan Beecher, ICPSR director of computing and network services. “With DuraCloud, behind the scenes we can replicate information and put it into different clouds, and there’s no difference to the user.” ICPSR has always stored multiple copies of its data, but Beecher feels that DuraCloud allows his institution to do so at lower cost, with less maintenance, and with fewer vendor relationships.

DuraCloud’s ability to scale as data grows is another plus to ICPSR, which--for the time being--will continue to utilize storage space at other institutions with which they have reciprocal relationships. These resources are not scalable, however, and hiccups--such as tripping the firewall--are fairly common. “If my stuff grows quickly, I can’t just ask for another terabyte of space," notes Beecher, who plans to end these one-off relationships as DuraCloud adds more providers. (ICPSR currently utilizes Amazon through DuraCloud.)

Even with a solution that allows clients to store data across multiple cloud providers, both Beecher and Kimpton are acutely aware that many academics will need time to fully embrace the concept of cloud storage. “There’s this knee-jerk reaction that you need to have physical control of content, but I’d say it’s safer in the cloud,” says Beecher, noting that just five years ago ICPSR managed a warehouse of data stored on magnetic tapes. "If an asteroid had hit Ann Arbor, everything would have been lost.”

According to Kimpton, the cautious approach taken by academics is understandable. "More than anyone else, cloud users tied to academic institutions tend to question the merits and risks of cloud storage and services," she notes. "They ask the hard questions. How will I know where my content is? How do I know it won’t go out of the US? What control do I have?" For Kimpton, these are all great questions, and she advises that everyone should ask them before signing on with a cloud provider.

Nevertheless, these questions all have answers, and Kimpton feels that it's important for colleges and universities not to use their concerns as an excuse to maintain the status quo. She calls the idea of technology no longer being controlled by IT departments “a revolution"--one that won’t be easy, in part because IT departments are all about control and mitigating risk. “In my view, cloud computing as SaaS is going to be a commodity, like electricity, where you’ll pay only for what you use. Adapting to this mindset--I think that's the next era."

Kimpton's hope is that a product offering multiple redundancies will ease worries about the cloud and smooth the transition for universities seeking secure archival solutions. To accommodate the various content-preservation needs of institutions, DuraSpace is offering three pricing structures for DuraCloud:

  • The DuraCloud Preservation Basic subscription stores two redundant copies of a university's original content at one cloud data center. It costs $1,500 per years for the first terabyte, $1,300 for additional terabytes.
  • The DuraCloud Preservation Plus plan stores four redundant copies of a university's original content at two cloud data centers. The cost is $3,000 a year for the first terabyte, $2,600 for additional terabytes.
  • The DuraCloud Enterprise subscription plan provides a full suite of configurable features for institutions that need multiple sub-accounts for departments, research groups, cross-institutional projects, or individuals. Depending on the number of cloud data centers used and the number of redundant copies, annual prices range from $5,900 to $7,200 for the first terabyte, and $1,300-$2,600 for additional terabytes.
comments powered by Disqus