Presentation 2001/7/12
Concurrency Control and Its Evaluation of the Fat-Btree
Jun MIYAZAKI, Haruo YOKOTA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We propose a concurrency control, named INC-OPT method, for the Fat-Btree. The Fat-Btree is designed as an index structure for shared-nothing parallel databases. The INC-OPT is a better concurrency control for parallel B-trees than conventional ones, because the INC-OPT primarily takes account of shared-nothing environments whereas the others were for a uni-processor or an SMP. In this paper, we give a practical result using an uCUBE3 machine that the INC-OPT enables the Fat-Btree to keep high throughput even under the high update ratio. Also, we show that the Fat-Btree outperforms other parallel B-trees such as an SIB and a CWB.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) parallel database / B-tree / concurrency control / performance evaluation / shared-nothing parallel computers
Paper # DE2001-83
Date of Issue

Conference Information
Committee DE
Conference Date 2001/7/12(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 JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Concurrency Control and Its Evaluation of the Fat-Btree
Sub Title (in English)
Keyword(1) parallel database
Keyword(2) B-tree
Keyword(3) concurrency control
Keyword(4) performance evaluation
Keyword(5) shared-nothing parallel computers
1st Author's Name Jun MIYAZAKI
1st Author's Affiliation School of Information Science, Japan Advanced Institute of Science and Technology()
2nd Author's Name Haruo YOKOTA
2nd Author's Affiliation Global Scientific Information and Computing Center, Tokyo Institute of Technology
Date 2001/7/12
Paper # DE2001-83
Volume (vol) vol.101
Number (no) 193
Page pp.pp.-
#Pages 8
Date of Issue