Dynamic Workload-Based Partitioning for Large-Scale Databases

Miguel Liroz-Gistau (1) , Reza Akbarinia (1) , Esther Pacitti (1) , Fabio Porto (2) , Patrick Valduriez (1)

1 ZENITH - Scientific Data Management
2 LNCC / MCT - Laboratorio Nacional de Computação Cientifica [Rio de Janeiro]

Applications with very large databases, where data items are continuously appended, are becoming more and more common. Thus, the development of efficient workload-based data partitioning is one of the main requirements to offer good performance to most of those applications that have complex access patterns, e.g. scientific applications. However, the existing workload-based approaches, which are executed in a static way, cannot be applied to very large databases. In this paper, we propose DynPart, a dynamic partitioning algorithm for continuously growing databases. DynPart efficiently adapts the data partitioning to the arrival of new data elements by taking into account the affinity of new data with queries and fragments. In contrast to existing static approaches, our approach offers a constant execution time, no matter the size of the database, while obtaining very good partitioning efficiency. We validated our solution through experimentation over real-world data; the results show its effectiveness.

Domaines: Base de données [cs.DB]; Calcul parallèle, distribué et partagé [cs.DC]

Miguel Liroz-Gistau, Reza Akbarinia, Esther Pacitti, Fabio Porto, Patrick Valduriez. Dynamic Workload-Based Partitioning for Large-Scale Databases. DEXA 2012 - 23rd International Conference on Database and Expert Systems Applications, Sep 2012, Vienna, Austria. pp.183-190, ⟨10.1007/978-3-642-32597-7_16⟩. ⟨lirmm-00748549⟩