Parallel Evaluation of Multi-join Queries

A.N. Wilschut; Jan Flokstra; Peter M.G. Apers

doi:10.1007/3-540-61695-0

Parallel Evaluation of Multi-join Queries

A.N. Wilschut, Jan Flokstra, Peter M.G. Apers

Databases (Former)

Research output: Contribution to conference › Paper › peer-review

182 Downloads (Pure)

Abstract

A number of execution strategies for parallel evaluation of multi-join queries have been proposed in the literature. In this paper we give a comparative performance evaluation of four execution strategies by implementing all of them on the same parallel database system, PRISMA/DB. Experiments have been done up to 80 processors. These strategies, coming from the literature, are named: Sequential Parallel, Synchronous Execution, Segmented Right-Deep, and Full Parallel. Based on the experiments clear guidelines are given when to use which strategy. This is an extended abstract; the full paper appeared in Proc. ACM SIGMOD'94, Minneapolis, Minnesota, May 24–27, 1994

Original language	Undefined
Pages	90-97
Number of pages	8
DOIs	https://doi.org/10.1007/3-540-61695-0
Publication status	Published - Sept 1996
Event	Third International ACPC Conference with Special Emphasis on Parallel Databases and Parallel I/O: Parallel Computation - Klagenfurt, Austria Duration: 23 Sept 1996 → 25 Sept 1996

Conference

Conference	Third International ACPC Conference with Special Emphasis on Parallel Databases and Parallel I/O: Parallel Computation
Period	23/09/96 → 25/09/96
Other	23-25 Sep 1996

Keywords

IR-66476
DB-PDB: PARALLEL DATABASES
EWI-7368

Access to Document

10.1007/3-540-61695-0

http://springerlink.metapress.com/content/f153p3uk64225m15

Cite this

@conference{42a2704185a34a6dad3cdce67fa043e6,

title = "Parallel Evaluation of Multi-join Queries",

abstract = "A number of execution strategies for parallel evaluation of multi-join queries have been proposed in the literature. In this paper we give a comparative performance evaluation of four execution strategies by implementing all of them on the same parallel database system, PRISMA/DB. Experiments have been done up to 80 processors. These strategies, coming from the literature, are named: Sequential Parallel, Synchronous Execution, Segmented Right-Deep, and Full Parallel. Based on the experiments clear guidelines are given when to use which strategy. This is an extended abstract; the full paper appeared in Proc. ACM SIGMOD'94, Minneapolis, Minnesota, May 24–27, 1994",

keywords = "IR-66476, DB-PDB: PARALLEL DATABASES, EWI-7368",

author = "A.N. Wilschut and Jan Flokstra and Apers, {Peter M.G.}",

note = "Imported from EWI/DB PMS [db-utwente:inpr:0000003073]; Third International ACPC Conference with Special Emphasis on Parallel Databases and Parallel I/O: Parallel Computation ; Conference date: 23-09-1996 Through 25-09-1996",

year = "1996",

month = sep,

doi = "10.1007/3-540-61695-0",

language = "Undefined",

pages = "90--97",

}

TY - CONF

T1 - Parallel Evaluation of Multi-join Queries

AU - Wilschut, A.N.

AU - Flokstra, Jan

AU - Apers, Peter M.G.

N1 - Imported from EWI/DB PMS [db-utwente:inpr:0000003073]

PY - 1996/9

Y1 - 1996/9

N2 - A number of execution strategies for parallel evaluation of multi-join queries have been proposed in the literature. In this paper we give a comparative performance evaluation of four execution strategies by implementing all of them on the same parallel database system, PRISMA/DB. Experiments have been done up to 80 processors. These strategies, coming from the literature, are named: Sequential Parallel, Synchronous Execution, Segmented Right-Deep, and Full Parallel. Based on the experiments clear guidelines are given when to use which strategy. This is an extended abstract; the full paper appeared in Proc. ACM SIGMOD'94, Minneapolis, Minnesota, May 24–27, 1994

AB - A number of execution strategies for parallel evaluation of multi-join queries have been proposed in the literature. In this paper we give a comparative performance evaluation of four execution strategies by implementing all of them on the same parallel database system, PRISMA/DB. Experiments have been done up to 80 processors. These strategies, coming from the literature, are named: Sequential Parallel, Synchronous Execution, Segmented Right-Deep, and Full Parallel. Based on the experiments clear guidelines are given when to use which strategy. This is an extended abstract; the full paper appeared in Proc. ACM SIGMOD'94, Minneapolis, Minnesota, May 24–27, 1994

KW - IR-66476

KW - DB-PDB: PARALLEL DATABASES

KW - EWI-7368

U2 - 10.1007/3-540-61695-0

DO - 10.1007/3-540-61695-0

M3 - Paper

SP - 90

EP - 97

T2 - Third International ACPC Conference with Special Emphasis on Parallel Databases and Parallel I/O: Parallel Computation

Y2 - 23 September 1996 through 25 September 1996

ER -