Result-set management for NDP operations on smart storage
- Current data-intensive systems suffer from scalability as they transfer massive amounts of data to the host DBMS to process it there. Novel near-data processing (NDP) DBMS architectures and smart storage can provably reduce the impact of raw data movement. However, transferring the result-set of an NDP operation may increase the data movement, and thus, the performance overhead. In this paper, we introduce a set of in-situ NDP result-set management techniques, such as spilling, materialization, and reuse. Our evaluation indicates a performance improvement of 1.13 × to 400 ×.
Author of HS Reutlingen | Petrov, Ilia |
---|---|
DOI: | https://doi.org/10.1145/3533737.3535097 |
ISBN: | 978-1-4503-9378-2 |
Erschienen in: | DaMoN '22 : proceedings of the 18th International Workshop on Data Management on New Hardware, Philadelphia, PA, June 2022 |
Publisher: | Association for Computing Machinery |
Place of publication: | New York |
Document Type: | Conference proceeding |
Language: | English |
Publication year: | 2022 |
Page Number: | 5 |
First Page: | 1 |
Last Page: | 5 |
Article Number: | 12 |
DDC classes: | 004 Informatik |
Open access?: | Nein |
Licence (German): | In Copyright - Urheberrechtlich geschützt |