Building the EJournal Safety Net: UConn's LOCKSS Experience

1
Building the EJournal Safety Net UConn's LOCKSS
Experience

• ACRL New England Chapter
• Information Technology Interest Group
• Serials Librarians Interest Group
• June 2, 2005

Dave Bretthauer Network Services
Librarian Information Technology Services/Network
Services Team University of Connecticut Libraries
2
Agenda

• UConn Libraries University
• Concerns About Perpetual Access
• UConns LOCKSS Involvement
• Cache Configuration
• Technical and Maintenance Issues
• Discoveries
• Questions
• New Directions
• Potential Challenges
• Learning Opportunities

3
University Libraries

• Librarians have flexibility in collection
  development not seen in some research
  institutions
• Network Services Team since 1997 made collection
  development decisions for large purchases and
  interdisciplinary electronic resources
• Permanent members are Information Technology
  Services and Collection Services staff
• Non-permanent members from Research and
  Information Services and Regional Campus
  Libraries (administratively part of Storrs)
• Law and Health Center Librarians also attend our
  meetings and are on the Teams listserv
• Current tools SerialsSolutions AMS, and MARC
  records LinkFinderPlus
• Licensing currently done by Acquisitions/Serials
  Team Leader

4
University Environment

• Remarkable Transformation
• Stress on metrics
• Shelf space dwindling

5
Libraries Response

• Dramatic jump in e-resource purchases
• Many deals via consortia, particularly NERL
• Cooperative collection development with Law
  School Library and Health Center Library
• Current structure is overloaded

6
Preservation Perpetual Access

• Awareness of current risk
• Ejournal preservation is too important not to be
  solved
• UConn willing to support ejournal preservation
  efforts
• In the long term, publishers and societies
  shouldnt have to bear the burden of maintaining
  availability of past output anymore than they did
  in the print world
• Perpetual access needs dependent on material

7
LOCKSS Involvement

• LOCKSS has been the only viable local ejournal
  preservation effort to which we can contribute
• Continues ownership model we had with print
• Annual cost for LOCKSS tiny fraction of our
  e-resource budget
• LOCKSS cache setup was an ITS goal for 2003-2004
• Set up cache shortly before LOCKSS moved to
  production, February 2004
• Cached journal/years as they became available for
  caching, including the LOCKSS Card for
  demonstration
• Joined LOCKSS Alliance in January 2005

8
Cache Installation/Configuration

• Current LOCKSS Box Dell Optiplex GX300, shipped
  10/2000
• Requested configuration information needed per
  LOCKSS instructions
• Added 120 GB Drive
• Located cache in librarys server farm, only
  because we have one

9
Technical/Maintenance Issues

• Maintenance is minimal
• adding to cache a few minutes per month
• updating CD perhaps 2-3 hours about every eight
  months when new CD releases are made
• Updating to a new CD requires physical access to
  the box
• Other done remotely via web interface
• UNIX/BSD experience not required to set up or
  maintain LOCKSS cache!

10
Discoveries

• Licensing
• To date we have relied on LOCKSS
• NERL is developing language for LOCKSS which we
  will adopt
• Others
• Cache disk space per journal/year is considerably
  less than we first expected
• Caching journal-years of subscription titles is
  time-sensitive
• Collection development discussion to date Cache
  what you can, if disk space becomes an issue
  well revisit this decision.
• Current lack of staff time to develop plugins

11
Questions

• Do we need to cache everything we might want to
  access later? Yes.
• What happens when we need to cache a large
  collection? Stop thinking 80GB is a large hard
  drive.
• How is the cached content accessible to our
  users? (LOCKSS Card example)
• How will we manage access to cached materials as
  we move away from IP authentication?
• Now that we have the cache, how do we real-world
  test it before we need it?

12
New DirectionsNot Just for EJournals Anymore!

• GPO LOCKSS Pilot Project
• Locally created digital collections
• Perpetual access?

13
Potential Challenges For Libraries Considering
LOCKSS

• Based on our experience libraries without
  internal IT departments may have difficulty
  getting permission to set up a cache from their
  campus IT centers
• Allowing LOCKSS Team cache access
• Stupid passwords are a problem for security, just
  as with any networked computer
• Many discussions about LOCKSS still need to take
  place

14
Learning Opportunities

• Slow organizational learning curve
• LOCKSS is really not an IT project
• Alliance membership decision

15
Thank you!

• Dave Bretthauer
• Network Services Librarian
• Information Technology Services/Network Services
  Team
• University of Connecticut Libraries