SARD: A statistical approach for ranking database tuning parameters

Biplob K. Debnath, David J. Lilja, Mohamed Mokbel

Research output: Chapter in Book/Report/Conference proceedingConference contribution

21 Citations (Scopus)

Abstract

Traditionally, DBMSs are shipped with hundreds of configuration parameters. Since the database performance highly depends on the appropriate settings of the configuration parameters, DBAs spend a lot of their time and effort to find the best parameter values for tuning the performance of the application of interest. In many cases, they rely on their experience and some rules of thumbs. However, time and effort may be wasted by tuning those parameters which may have no or marginal effects. Moreover, tuning effects also vary depending on the expertise of the DBAs, but skilled DBAs are increasingly becoming rare and expensive to employ. To address these problems, we present a Statistical Approach for Ranking Database parameters (SARD), which is based on the Plackett & Burman statistical design methodology. SARD takes the query workload and the number of configuration parameters as inputs, and using only a linear number of experiments, generates a ranking of database parameters based on their relative impacts on the DBMS performance. Preliminary experimental results using TPC-H and PostgreSQL show that SARD generated ranking can correctly identify critical configuration parameters.

Original languageEnglish
Title of host publicationProceedings of the 2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08
Pages11-18
Number of pages8
DOIs
Publication statusPublished - 1 Sep 2008
Externally publishedYes
Event2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08 - Cancun, Mexico
Duration: 7 Apr 200812 Apr 2008

Other

Other2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08
CountryMexico
CityCancun
Period7/4/0812/4/08

Fingerprint

Tuning
Experiments

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Information Systems

Cite this

Debnath, B. K., Lilja, D. J., & Mokbel, M. (2008). SARD: A statistical approach for ranking database tuning parameters. In Proceedings of the 2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08 (pp. 11-18). [4498279] https://doi.org/10.1109/ICDEW.2008.4498279

SARD : A statistical approach for ranking database tuning parameters. / Debnath, Biplob K.; Lilja, David J.; Mokbel, Mohamed.

Proceedings of the 2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08. 2008. p. 11-18 4498279.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Debnath, BK, Lilja, DJ & Mokbel, M 2008, SARD: A statistical approach for ranking database tuning parameters. in Proceedings of the 2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08., 4498279, pp. 11-18, 2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08, Cancun, Mexico, 7/4/08. https://doi.org/10.1109/ICDEW.2008.4498279
Debnath BK, Lilja DJ, Mokbel M. SARD: A statistical approach for ranking database tuning parameters. In Proceedings of the 2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08. 2008. p. 11-18. 4498279 https://doi.org/10.1109/ICDEW.2008.4498279
Debnath, Biplob K. ; Lilja, David J. ; Mokbel, Mohamed. / SARD : A statistical approach for ranking database tuning parameters. Proceedings of the 2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08. 2008. pp. 11-18
@inproceedings{f21ccae849ef409591793d35199e7caa,
title = "SARD: A statistical approach for ranking database tuning parameters",
abstract = "Traditionally, DBMSs are shipped with hundreds of configuration parameters. Since the database performance highly depends on the appropriate settings of the configuration parameters, DBAs spend a lot of their time and effort to find the best parameter values for tuning the performance of the application of interest. In many cases, they rely on their experience and some rules of thumbs. However, time and effort may be wasted by tuning those parameters which may have no or marginal effects. Moreover, tuning effects also vary depending on the expertise of the DBAs, but skilled DBAs are increasingly becoming rare and expensive to employ. To address these problems, we present a Statistical Approach for Ranking Database parameters (SARD), which is based on the Plackett & Burman statistical design methodology. SARD takes the query workload and the number of configuration parameters as inputs, and using only a linear number of experiments, generates a ranking of database parameters based on their relative impacts on the DBMS performance. Preliminary experimental results using TPC-H and PostgreSQL show that SARD generated ranking can correctly identify critical configuration parameters.",
author = "Debnath, {Biplob K.} and Lilja, {David J.} and Mohamed Mokbel",
year = "2008",
month = "9",
day = "1",
doi = "10.1109/ICDEW.2008.4498279",
language = "English",
isbn = "9781424421626",
pages = "11--18",
booktitle = "Proceedings of the 2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08",

}

TY - GEN

T1 - SARD

T2 - A statistical approach for ranking database tuning parameters

AU - Debnath, Biplob K.

AU - Lilja, David J.

AU - Mokbel, Mohamed

PY - 2008/9/1

Y1 - 2008/9/1

N2 - Traditionally, DBMSs are shipped with hundreds of configuration parameters. Since the database performance highly depends on the appropriate settings of the configuration parameters, DBAs spend a lot of their time and effort to find the best parameter values for tuning the performance of the application of interest. In many cases, they rely on their experience and some rules of thumbs. However, time and effort may be wasted by tuning those parameters which may have no or marginal effects. Moreover, tuning effects also vary depending on the expertise of the DBAs, but skilled DBAs are increasingly becoming rare and expensive to employ. To address these problems, we present a Statistical Approach for Ranking Database parameters (SARD), which is based on the Plackett & Burman statistical design methodology. SARD takes the query workload and the number of configuration parameters as inputs, and using only a linear number of experiments, generates a ranking of database parameters based on their relative impacts on the DBMS performance. Preliminary experimental results using TPC-H and PostgreSQL show that SARD generated ranking can correctly identify critical configuration parameters.

AB - Traditionally, DBMSs are shipped with hundreds of configuration parameters. Since the database performance highly depends on the appropriate settings of the configuration parameters, DBAs spend a lot of their time and effort to find the best parameter values for tuning the performance of the application of interest. In many cases, they rely on their experience and some rules of thumbs. However, time and effort may be wasted by tuning those parameters which may have no or marginal effects. Moreover, tuning effects also vary depending on the expertise of the DBAs, but skilled DBAs are increasingly becoming rare and expensive to employ. To address these problems, we present a Statistical Approach for Ranking Database parameters (SARD), which is based on the Plackett & Burman statistical design methodology. SARD takes the query workload and the number of configuration parameters as inputs, and using only a linear number of experiments, generates a ranking of database parameters based on their relative impacts on the DBMS performance. Preliminary experimental results using TPC-H and PostgreSQL show that SARD generated ranking can correctly identify critical configuration parameters.

UR - http://www.scopus.com/inward/record.url?scp=50249173350&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=50249173350&partnerID=8YFLogxK

U2 - 10.1109/ICDEW.2008.4498279

DO - 10.1109/ICDEW.2008.4498279

M3 - Conference contribution

AN - SCOPUS:50249173350

SN - 9781424421626

SP - 11

EP - 18

BT - Proceedings of the 2008 - IEEE 24th International Conference on Data Engineering Workshop, ICDE'08

ER -