DSpace Repository Planning Guide

Start with the repository’s institutional purpose

Define what the repository will hold, who may deposit, who reviews submissions, what becomes public, and which long-term responsibilities the institution accepts. Theses, faculty publications, datasets, heritage objects, and administrative records may require different metadata, access, and preservation decisions.

Repository planning checklist

• Map communities, collections, owners, depositors, reviewers, administrators, and public audiences.
• Select metadata fields, controlled vocabularies, identifiers, languages, rights statements, and required values.
• Design submission, review, correction, approval, withdrawal, embargo, and publication workflows.
• Plan discovery through search, browse, facets, sitemaps, indexing, persistent links, and external harvesters.
• Document storage, database, asset-store, configuration, backup, monitoring, and restoration responsibilities.

Metadata is a service decision

Metadata should support discovery, reporting, interoperability, and future migration. Agree field definitions and examples before bulk ingest. Validate creators, dates, types, subjects, departments, degrees, rights, and identifiers consistently; avoid collecting fields no one can maintain.

Design workflows people can operate

Keep roles and approval steps proportionate to the institution. Test ordinary submissions and difficult cases such as corrections, embargoes, large files, replacement versions, restricted material, and withdrawn items. Administrators should know how to diagnose an item that is submitted but not discoverable.

Plan for preservation and recovery

A repository is not preserved simply because files are online. Maintain verified backups of the database, asset store, configuration, and customizations; keep copies outside the production host; monitor storage and scheduled tasks; and rehearse restoration. Record format, rights, retention, and integrity decisions as the collection grows.