InVerDa – The Liquid Database

Abstract

Multiple applications, which share one common database, will evolve over time by their very nature. Often, former versions need to stay available, so database developers find themselves maintaining co-existing schema versions of multiple applications in multiple versions—usually with handwritten delta code—which is highly error-prone and explains significant costs in software projects. We showcase I V D , a tool using the richer semantics of a bidirectional database evolution language to generate all the delta code automatically, easily providing co-existing schema versions within one database. I V D automatically decides on an optimized physical database schema serving all schema versions to transparently optimize the performance for the current workload.