Data as Technical Debt: Managing the Challenges of Legacy Data During a System Migration

Abstract

Software developers use the term “technical debt” to refer to elements of software that, while effective in the short term, prove expensive or unsustainable in the long run, requiring a special effort to upgrade, migrate, and/or maintain. Adapting the term to the management of integrated library systems, this presentation uses “technical debt” to describe challenges posed by legacy data upon migration to a new system. In our case, migration from Voyager to Alma exposed various forms of this debt: incurred as the result of constraints built into the Voyager environment; as a by-product of workflows geared toward transactional uses of the data but that prove ill-suited to needs for data in the aggregate; or because of a lack of continuity in processes and procedures over time. Alma's bulk jobs and API's present opportunities to manage this debt in a more holistic and efficient way, although other elements of Alma's design and of the migration process itself generate their own forms of technical debt (e.g., the impact on e-resource orders of the p2e process). This presentation will discuss concrete examples drawn from multiple functional areas, including acquisitions and e-resources, as well as strategies for correcting these problems (and their associated benefits and costs). By highlighting the ways in which technical debt creeps into our library systems, we hope to promote a focus on workflows that can reduce such debt in the future, ensuring the long-term sustainability of our data.