Presentation 1997/12/3
Parallel Query Execution Method on a Large Scale PC Cluster and Its Evaluation
Takayuki Tamura, Masato Oguchi, Masaru Kitsuregawa,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We developed a PC cluster system which consists of 100 PCs as a test bed for massively parallel query processing. Each PC employs the 200MHz Pentium Pro CPU and is connected with others through an ATM switch. On each PC node, a database kernel is running to process the queries in cooperation with other nodes. The kernel supports efficient execution of pipelined operators, which is crucial to the high performance on complex decision support queries. The performance evaluation result with the standard benchmark, TPC-D, on a 100 GB database was quite good compared with commercial parallel systems. In this paper, we explore two techniques which contribute the performance improvement. First, we applied transposed file organization, which vertically partitions the relations to allow attribute-by-attribute access. This resulted in significant performance improvement by reducing the amount of disk I/O. Second, to absorb the extra CPU costs incurred by this technique, we implemented a execution method where multiple operators were executed within a single thread to eliminate the overhead of context switches.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) PC Cluster / Parallel Query Processing / Transposed Files / Context Switch
Paper # DE97-88
Date of Issue

Conference Information
Committee DE
Conference Date 1997/12/3(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Data Engineering (DE)
Language ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Parallel Query Execution Method on a Large Scale PC Cluster and Its Evaluation
Sub Title (in English)
Keyword(1) PC Cluster
Keyword(2) Parallel Query Processing
Keyword(3) Transposed Files
Keyword(4) Context Switch
1st Author's Name Takayuki Tamura
1st Author's Affiliation Institute of Industrial Science, The University of Tokyo()
2nd Author's Name Masato Oguchi
2nd Author's Affiliation Institute of Industrial Science, The University of Tokyo
3rd Author's Name Masaru Kitsuregawa
3rd Author's Affiliation Institute of Industrial Science, The University of Tokyo
Date 1997/12/3
Paper # DE97-88
Volume (vol) vol.97
Number (no) 418
Page pp.pp.-
#Pages 6
Date of Issue