LOCKSS (Lots of Copies Keeps Stuff Safe)

Taking its name (Lots of Copies Keeps Stuff Safe) from the idea that a file is more likely to survive if it has multiple copies stored in multiple places, LOCKSS software allows libraries to create preserved digital collections out of materials that would otherwise be accessible only through a licensed academic subscription. LOCKSS has agreements with academic publishers to permit libraries to store local copies of content that is usually only accessible through web-based subscriptions, enabling post-cancellation content access. Serving as a web proxy, the software harvests content directly from publisher websites and continually compares it with the same content collected by other libraries’ systems, so that repairs can be made if any differences are detected.

The computer serving as a “LOCKSS Box” becomes a digital preservation appliance, allowing staff to monitor their content, control access, and perform file format migration. In this capacity they can be used in Private LOCKSS Networks (PLNs), working with material supplied by the organisations themselves.

Provider

LOCKSS is a stand-alone program based at Stanford University Libraries. The JISC-endorsed UK LOCKSS Alliance provides support for UK institutions.

Licensing and cost

The software is free and is available under an open-source Stanford-written license. However, US institutions wishing to join the LOCKSS Alliance pay an annual fee ranging from $2,236 for “Baccalaureate Colleges” to $11,180 for “Very Large Research Universities”. For UK institutions, the UK LOCKSS Alliance offers pro-rata membership from £1,800-5,000 per year.

Development activity

The LOCKSS program is currently self-sustaining, and so development is ongoing. The software is updated on a six-week cycle, and automatically pushed to LOCKSS Boxes.

Platform and interoperability

LOCKSS software is integrated with a Linux installation based on CentOS 5. They therefore recommend installing LOCKSS on a dedicated workstation or server (physical or virtual), which should ideally be configured with 2GB of RAM and two 1TB hard drives to support RAID disk redundancy. LOCKSS was previously provided on an OpenBSD based platform, but as of 2011 this platform is being phased out.

LOCKSS includes a Java framework to support plugin writers.

Functional notes

The LOCKSS system requires certain publisher-specific knowledge and code for each publication. This information is contained in a “plugin” module, which can be written and supplied by publishers, the community, or LOCKSS staff, and is then pushed out to libraries subscribing to those journals.

Libraries must have active subscriptions to collect content; as of 2009 the average replication factor in the public LOCKSS network is approximately 40.

Users can set up automatic Linux upgrades.

Documentation and user support

LOCKSS has excellent documentation through its website, including an Installation Guide, and an OpenBSD to Linux Conversion Guide. LOCKSS Alliance members receive technical and collections support from the Stanford University LOCKSS team; non-members to not receive direct user support.

Usability

Installation and configuration are non-trivial, but LOCKSS offers a customised Linux installation CD (the LOCKSS Net Install CD) that details the process.

The administrative interface is an extremely simple GUI, and content access occurs through the institution’s current web interface.

Expertise required

Installation and maintenance require solid knowledge of systems administration, particularly for Linux. Administratively, users must have sound collections policies and familiarity with licensing agreements.

Standards compliance

LOCKSS advertises itself as providing an OAIS-compliant infrastructure.

Influence and take-up

As of 2010, the LOCKSS Alliance had approximately 200 members. However, the software has had nearly 3900 downloads from the Sourceforge site.

In addition, the website lists a number of examples of organisations housing Private LOCKSS Networks (PLNs), including Data PASS (Data Preservation Alliance for the Social Sciences), Digital Commons of Berkeley Electronic Press, and the Alabama Digital Preservation Network.

Last reviewed: 
8 July, 2013