The aim of project is to enhance the content of repositories throughout the UK's HE and FE sector by providing solutions for the scalability of repositories as they grow in size and complexity. As a rapid innovation project, it looks to remove the "hardware" barriers involved in launching and maintaining a repository. It will accomplish this initially by enabling the use of Fedora Commons repository software on-top-of Amazon's virtual servers (EC2 & S3). By pre-configuring these servers, any HE/FE institution can "rent" Amazon server space and launch their own secure Fedora repository without having to pre-configure a local server within their institution. In short, institutions can launch their repository service in the same day they decide to have one, and without hiring a "hardware" expert. Overall, the project will begin to formulate the cost effectiveness for this kind of set-up and recommend best-practice to other repository departments.

Fedorazon

See the final report

Rapid Innovation Project

The Aim of project Fedorazon is to enhance the content of repositories throughout the UK's HE and FE sector by providing solutions for the scalability of repositories as they grow in size and complexity. As a rapid innovation project, it looks to remove the "hardware" barriers involved in launching and maintaining a repository. It will accomplish this initially by enabling the use of Fedora Commons repository software on-top-of Amazon's virtual servers (EC2 & S3). By pre-configuring these servers, any HE/FE institution can "rent" Amazon server space and launch their own secure Fedora repository without having to pre-configure a local server within their institution. In short, institutions can launch their repository service in the same day they decide to have one, and without hiring a "hardware" expert.  Overall, the project will begin to formulate the cost effectiveness for this kind of set-up and recommend best-practice to other repository departments. 

Overview

One of the primary barriers that have been encountered across the UK HE and FE repository community is the lack of technical knowledge in the departments that are most apt to take on repositories.  A specific user case -that is often cited- is the willingness by libraries to take on the fundamental role of digital content management, but do not have the hardware nor expertise of how that hardware can scale and maintain its robustness.  In addition, where some libraries might have the technical know-how, they do not necessarily have the facilities to host multiple physical servers nor the budget to do so (The Bloomsbury College Libraries are particularly aware of this problem given the cost of hosting local servers in WC1 space).  This fundamental hardware computing requirement must be addressed if the 14 million pounds that JISC has invested is not to be drained by the ever increasing need for more servers as more content is collected.  In other words, the complexity of physical hardware set-up and maintenance must be hidden away from the institution so true focus can be placed upon enhancing repository content.

Aims and objectives 

Aim: To improve the overall capabilities of hardware configuration for repository installations.

  • Provide OS code and “how-to” documentation for using Fedora with EC2, S3 and other services
  • Encourage and disseminate the use of Fedora on AWS to JISC projects and communities
  • Provide training on how to deploy Fedora as a component of other services
  • Support any and all project wishing to deploy repository services on EC2 and S3
  • Provide robust testing and validation of these services as part of the e-Framework & JISC SOA communities.

Project methodology

The project is based at Birkbeck Library where the head of the library (Philip Payne) oversees new developments in library technology.  The project manager is based in the library systems team in Bloomsbury where the resources and technical support for the project are ample with six college partners supporting this work.  The project manager (David F. Flanders) –with experience in several other JISC and HEFCE projects- will oversee this project with 20% of his time.  He will oversee the project and produce the majority of reports and research documentation.  He will also oversee the assignment of any technological workpackages that will be assigned to MediaShelf (Matt Zumalt) and King’s College (Mark Hedges).  Technical documentation will be edited by the project manager, but in the first instance written by the technology team member undertaking the work.

Anticipated outputs and outcomes

The best possible outcome of the project would be the ability by JISC institutions to forego the need to have technical administration for hardware servers.  This would leave repository staff to focus on the real work of the repository: content population and enhancement.  However, the overall outcome of this project needs be the education of repository managers in regards to their options for hardware for their repository (knowledge and experience deliverables above).  Once repositories do become the de facto place to put institutional content the requirements by the repository to scale robustly is paramount.

Technology / Standards used

HTTP, ReST API (Fedora), ReST API (S3), EC2 AMI, WebDAV, DavFS, SSH, SSL, MySQL 5.x, Pound, Tomcat

Project Staff

Project Manager

Documents & Multimedia

Bookmark and Share
Summary
Start date
1 October 2007
End date
31 March 2008
Funding programme
Repositories and Preservation Programme
Strand
Repositories Start-up and Enhancement projects
Project website
Lead institutions
Birkbeck College
Partner institutions
  • London School of Hygeine and Tropical Medicine
  • Royal Veterinary College
  • School of Oriental and African Studies
  • School of Pharmacy, University of London
  • University College London
Committees
  • JISC Integrated Information Environment committee
Topic