Metric and Workload Effects on Computer Systems Evaluation

Dror G. Feitelson

doi:10.1109/MC.2003.1231190

Metric and Workload Effects on Computer Systems Evaluation

Dror G. Feitelson^*

^*Corresponding author for this work

The Rachel and Selim Benin School of Engineering and Computer Science

Research output: Contribution to specialist publication › Article

50 Scopus citations

Abstract

While a system's performance is a function of its design and implementation the workload researchers subject the system to and the metrics they use also can affect performance evaluation results. There are two main approaches to performance evaluation: Analysis involves simplifications in the interest of mathematical tractability, and simulation offers a more realistic alternative that directly uses real workload recordings. The problems the author describes appear more frequently during simulation because of its ability to directly reflect complex situations, even those the evaluator does not know about or understand.

Original language	English
Pages	18-25+4
Volume	36
No	9
Specialist publication	Computer
DOIs	https://doi.org/10.1109/MC.2003.1231190
State	Published - Sep 2003

Bibliographical note

Funding Information:
The Israel Science Foundation, grant no. 219/99, supported this research in part. The workload models and logs on which they are based can be obtained online from the Parallel Workloads Archive. Thanks to all those who made their data and software available.

Access to Document

10.1109/MC.2003.1231190

Cite this

@misc{17350dcb4d67482eb2d0b29d845a4b22,

title = "Metric and Workload Effects on Computer Systems Evaluation",

abstract = "While a system's performance is a function of its design and implementation the workload researchers subject the system to and the metrics they use also can affect performance evaluation results. There are two main approaches to performance evaluation: Analysis involves simplifications in the interest of mathematical tractability, and simulation offers a more realistic alternative that directly uses real workload recordings. The problems the author describes appear more frequently during simulation because of its ability to directly reflect complex situations, even those the evaluator does not know about or understand.",

author = "Feitelson, {Dror G.}",

note = "Funding Information: The Israel Science Foundation, grant no. 219/99, supported this research in part. The workload models and logs on which they are based can be obtained online from the Parallel Workloads Archive. Thanks to all those who made their data and software available.",

year = "2003",

month = sep,

doi = "10.1109/MC.2003.1231190",

language = "אנגלית",

volume = "36",

pages = "18--25+4",

journal = "Computer",

issn = "0018-9162",

publisher = "IEEE Computer Society",

}

TY - GEN

T1 - Metric and Workload Effects on Computer Systems Evaluation

AU - Feitelson, Dror G.

N1 - Funding Information: The Israel Science Foundation, grant no. 219/99, supported this research in part. The workload models and logs on which they are based can be obtained online from the Parallel Workloads Archive. Thanks to all those who made their data and software available.

PY - 2003/9

Y1 - 2003/9

N2 - While a system's performance is a function of its design and implementation the workload researchers subject the system to and the metrics they use also can affect performance evaluation results. There are two main approaches to performance evaluation: Analysis involves simplifications in the interest of mathematical tractability, and simulation offers a more realistic alternative that directly uses real workload recordings. The problems the author describes appear more frequently during simulation because of its ability to directly reflect complex situations, even those the evaluator does not know about or understand.

AB - While a system's performance is a function of its design and implementation the workload researchers subject the system to and the metrics they use also can affect performance evaluation results. There are two main approaches to performance evaluation: Analysis involves simplifications in the interest of mathematical tractability, and simulation offers a more realistic alternative that directly uses real workload recordings. The problems the author describes appear more frequently during simulation because of its ability to directly reflect complex situations, even those the evaluator does not know about or understand.

UR - http://www.scopus.com/inward/record.url?scp=0141605916&partnerID=8YFLogxK

U2 - 10.1109/MC.2003.1231190

DO - 10.1109/MC.2003.1231190

M3 - מאמר בכתב עת

AN - SCOPUS:0141605916

SN - 0018-9162

VL - 36

SP - 18-25+4

JO - Computer

JF - Computer

ER -

Metric and Workload Effects on Computer Systems Evaluation

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this