A Row-column Hybrid Storage Approach based on Massively Parallel Processing

DOI

10.2991/icwcsn-16.2017.62How to use a DOI?

Keywords

massively parallel processing; row-column hybrid Storage; cross-layer query; relational database.

Abstract

Massively parallel processing architecture, as a common parallel processing server architecture, is widely used in the field of mathematical modeling and database processing which require a lot of computing. This architecture can complete the calls of multiple processors by the coordination, which can process the same command simultaneously. However, this architecture is unable to satisfy the needs of optimizing and time delay when deal with the relational database system of row or column storage. This paper presents a row-column hybrid storage approach based on massively parallel processing, which selects the method by judging the requirements of the pending data. The optimized row storage method is selected with the low time-delay requirement and the optimized column storage method is selected with the complex processing requirement. And with the help of cross-layer during query and processing, both on-line analytical processing and on-line transaction processing can be achieved in one system. The real experiments on massively parallel processing network architecture show that this method can achieve the balance between high throughput and complex processing, and improve the overall performance.

Copyright

© 2017, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Qiao Sun
AU  - Lan-Mei Fu
AU  - Peng Wu
AU  - Jia-Song Sun
PY  - 2016/12
DA  - 2016/12
TI  - A Row-column Hybrid Storage Approach based on Massively Parallel Processing
BT  - Proceedings of the 3rd International Conference on Wireless Communication and Sensor Networks (WCSN 2016)
PB  - Atlantis Press
SP  - 284
EP  - 288
SN  - 2352-538X
UR  - https://doi.org/10.2991/icwcsn-16.2017.62
DO  - 10.2991/icwcsn-16.2017.62
ID  - Sun2016/12
ER  -