From PreservWiki

Jump to: navigation, search

\title{Autonomous Preservation Services using Smart Storage}


{David Tarrnat \and Steve Hitchcock \\ School of Electronics and Computer Science \\ University of Southampton \\ Southampton \\ UK \\ SO17 1BJ \\ dct05r,sh94r @ecs.soton.ac.uk \And

Adrian Brown \\ The National Archives \\ Kew, Richmond \\Surrey \\ UK \\ TW9 4DU \\ adrian.brown@nationalarchives.gov.uk}




  • COMMENT: Similar to PASIG one but more focused on the framework of using smart storage to enable autonomous preservation. However the repository software has to be on board to help manage the process and supply policy.

EPrints, the first software for building institutional repositories, has recently been focussing on user-centred developments (ref. Carr, PASIG, Paris, 2007), for example, improving metadata quality and enabling easier and faster repository customisation. Future releases will incorporate data-oriented developments. Working with the UK-based Preserv 2 project, the next version will introduce a storage controller that supports a pluggable storage layer for repositories, providing the ability to store objects in different locations based on metadata or type. The storage controller enables direct interaction between the repository software and open storage platforms. Building on this we envisage platforms that not only store digital objects and associated metadata, but also facilitate services to act on these objects for long-term preservation. We refer to this concept as smart storage. This talk will explain how interoperability has been recognised by the Preserv 2 project as a key component of repository software to enable interaction with preservation services. The methods EPrints is employing to support both open storage and smart storage platforms will be outlined, showing how these contribute to long-term digital preservation.


\section{Page 1}

Here we want to explain the background of preserv and introduce both Open Storage and Smart Storage. We firstly want to outline the structure of preserv and include my famous diagram with the various layers and clouds. We can then explain how preservation services and digital repositories currently have a technical gap between them where on is unable to directly interact with the other in a controlled fashion, and how un until now harvesting has been the key way to get items from one repository into another. We can then outline open storage and how having a storage controller as part of a repository is a very impporant enabled not just to complete preservation of a digital object but also enables us to interface with smart storage. We can then explain what smart storage is and why it is so important and what services it can provide. We can then outline that preserv is looking to contructing a framework of services which can be run in many ways including on a smart storage platform which then feeds back information to the repository or used to interface directly with the repository and it's objects. We can then lead into the example of the DROID tool.

\section{Page 2}

Here we outline the DROID tool and it's current functionality. We can then outline the early problems with the tool and state how by being able to move it between being an invoked web service, and an autonomous service in the repository software or smart storage enables us to demonstate when more services of this type can be run.

We also need to explain the importance of policy with some tools as each repository will weight some tools stronger than others, we can give the example of significant properties weightings here. If the Smart Storage is going to autonomously migrate formats it needs to be able to do this in a manor which suits the repository policy (but this may not be the only migration which is performed).

We can then conclude by outlining how other services could be run at the various levels to enable preservation through automatic migration and how it is inportant to keep the repository manager in control of their own content but not necessarily that of the smart store.




Notes for Steve

  • What and how much background on Preserv do we need in this abstract (references)
  • Is section 1 (page 1) suitable or is it too deep and thus unfocused.
  • What more needs to go in Adrians section (page 2), i'm still getting to grips with DROID and PRONOM.
Personal tools