Scalable transaction processing through data-oriented execution

Data management technology changes the world we live in by providing efficient access to huge volumes of constantly changing data and by enabling sophisticated analysis of those data. In parallel, we witness a tremendous shift in the underlying hardware technology toward highly parallel multicore processors. Data management systems need to fully exploit the abundantly available hardware parallelism. Transaction processing is one of the most important and challenging database workloads and this dissertation contributes to the quest for scalable transaction processing software. It shows that conventional transaction processing has inherent scalability limitations due to the unpredictable access patterns caused by the request-oriented execution model it follows. Instead, it proposes adopting a data-oriented execution model, and shows that transaction processing systems designed around data-oriented execution break the inherent limitations of conventional execution. The data-oriented design paves the way for transaction processing systems to maintain scalability as parallelism increases for the foreseeable future; as hardware parallelism increases, the benefits will only increase.

Autorentext

Ippokratis Pandis is a member of the research staff at IBM Research - Almaden. Prior joining IBM, he received his Ph.D. from the Electrical and Computer Engineering department of Carnegie Mellon University. His research focuses on high-performing and scalable data management on emerging hardware.