Building Flexible & Replicable Digital Repositories

Abstract

Drawing on prior successes with modular digital repositories, The University of Texas at Austin Libraries is continuing to evolve our architecture and thought processes as we embark on a series of strategic projects. Moving away from monolithic frameworks run on a virtual environment, we are standardizing development of new digital asset management systems and repositories with the use of Dockerized container-based services deployed with Kubernetes and using Github Actions to create a Continuous Integration, Continuous Delivery (CI/CD) pipeline. This approach significantly improves ease of deployment, scalability, resource management, maintenance, reproducibility, and future-proofing. Our new repositories are designed with modular and swappable services that allow for future iteration on or replacement of significant pieces without the need for a total rebuild of the underlying applications. Custom backend services providing python-based APIs and messaging queues provide the foundation for a diverse ecosystem that can accommodate an array of front end frameworks and other integrations. Our current initiatives include increased leverage of the Oxford Common File Layout (OCFL) via Fedora 6 in combination with SOLR to minimize the less-reproducible customization of unique relational databases. We are also exploring solutions to batch ingestion with repository agnostic microservices that can offer greater flexibility, performance, and more efficient resource allocation. Our briefing will discuss the strategies, technologies, and frameworks we are utilizing as we continue to evolve how we design and build digital archives for a wide variety of collections.