Performance analysis using Subsuming Methods: an industrial case study

David Maplesden; Karl Von Randow; Ewan Tempero; John Hosking; John Grundy

doi:10.1109/ICSE.2015.143

Performance analysis using Subsuming Methods: an industrial case study

David Maplesden, Karl Von Randow, Ewan Tempero, John Hosking, John Grundy

Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review

8 Citations (Scopus)

Abstract

Large-scale object-oriented applications consist of tens of thousands of methods and exhibit highly complex runtime behaviour that is difficult to analyse for performance. Typical performance analysis approaches that aggregate performance measures in a method-centric manner result in thinly distributed costs and few easily identifiable optimisation opportunities. Subsuming methods analysis is a new approach that aggregates performance costs across repeated patterns of method calls that occur in the application's runtime behaviour. This allows automatic identification of patterns that are expensive and represent practical optimisation opportunities. To evaluate the practicality of this analysis with a real world large-scale object-oriented application we completed a case study with the developers of letterboxd.com - a social network website for movie goers. Using the results of the analysis we were able to rapidly implement changes resulting in a 54.8% reduction in CPU load and an 49.6% reduction in average response time.

Original language	English
Title of host publication	Proceedings - 2015 IEEE/ACM 37th IEEE International Conference on Software Engineering, ICSE 2015, Volume 2
Editors	Sebastian Elbaum, Gerardo Canfora
Place of Publication	Piscataway NJ USA
Publisher	IEEE, Institute of Electrical and Electronics Engineers
Pages	149-158
Number of pages	10
Volume	2
ISBN (Electronic)	9781479919345
DOIs	https://doi.org/10.1109/ICSE.2015.143
Publication status	Published - 2015
Externally published	Yes
Event	International Conference on Software Engineering 2015 - Florence, Italy Duration: 16 May 2015 → 24 May 2015 Conference number: 37th http://2015.icse-conferences.org/ https://ieeexplore.ieee.org/xpl/conhome/7174815/proceeding?isnumber=7194545 (Proceedings)

Conference

Conference	International Conference on Software Engineering 2015
Abbreviated title	ICSE 2015
Country/Territory	Italy
City	Florence
Period	16/05/15 → 24/05/15
Internet address	http://2015.icse-conferences.org/ https://ieeexplore.ieee.org/xpl/conhome/7174815/proceeding?isnumber=7194545 (Proceedings)

Keywords

Object oriented software
Performance analysis
Runtime bloat
Subsuming methods

Access to Document

10.1109/ICSE.2015.143

Cite this

Maplesden, D., Randow, K. V., Tempero, E., Hosking, J., & Grundy, J. (2015). Performance analysis using Subsuming Methods: an industrial case study. In S. Elbaum, & G. Canfora (Eds.), Proceedings - 2015 IEEE/ACM 37th IEEE International Conference on Software Engineering, ICSE 2015, Volume 2 (Vol. 2, pp. 149-158). Article 7202959 IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/ICSE.2015.143

Maplesden, David ; Randow, Karl Von ; Tempero, Ewan et al. / Performance analysis using Subsuming Methods : an industrial case study. Proceedings - 2015 IEEE/ACM 37th IEEE International Conference on Software Engineering, ICSE 2015, Volume 2. editor / Sebastian Elbaum ; Gerardo Canfora. Vol. 2 Piscataway NJ USA : IEEE, Institute of Electrical and Electronics Engineers, 2015. pp. 149-158

@inproceedings{a7c6e0e6cb454983912fc777dd8c160f,

title = "Performance analysis using Subsuming Methods: an industrial case study",

abstract = "Large-scale object-oriented applications consist of tens of thousands of methods and exhibit highly complex runtime behaviour that is difficult to analyse for performance. Typical performance analysis approaches that aggregate performance measures in a method-centric manner result in thinly distributed costs and few easily identifiable optimisation opportunities. Subsuming methods analysis is a new approach that aggregates performance costs across repeated patterns of method calls that occur in the application's runtime behaviour. This allows automatic identification of patterns that are expensive and represent practical optimisation opportunities. To evaluate the practicality of this analysis with a real world large-scale object-oriented application we completed a case study with the developers of letterboxd.com - a social network website for movie goers. Using the results of the analysis we were able to rapidly implement changes resulting in a 54.8% reduction in CPU load and an 49.6% reduction in average response time.",

keywords = "Object oriented software, Performance analysis, Runtime bloat, Subsuming methods",

author = "David Maplesden and Randow, {Karl Von} and Ewan Tempero and John Hosking and John Grundy",

year = "2015",

doi = "10.1109/ICSE.2015.143",

language = "English",

volume = "2",

pages = "149--158",

editor = "Elbaum, {Sebastian } and Canfora, {Gerardo }",

booktitle = "Proceedings - 2015 IEEE/ACM 37th IEEE International Conference on Software Engineering, ICSE 2015, Volume 2",

publisher = "IEEE, Institute of Electrical and Electronics Engineers",

address = "United States of America",

note = "International Conference on Software Engineering 2015, ICSE 2015 ; Conference date: 16-05-2015 Through 24-05-2015",

url = "http://2015.icse-conferences.org/, https://ieeexplore.ieee.org/xpl/conhome/7174815/proceeding?isnumber=7194545",

}

Maplesden, D, Randow, KV, Tempero, E, Hosking, J & Grundy, J 2015, Performance analysis using Subsuming Methods: an industrial case study. in S Elbaum & G Canfora (eds), Proceedings - 2015 IEEE/ACM 37th IEEE International Conference on Software Engineering, ICSE 2015, Volume 2. vol. 2, 7202959, IEEE, Institute of Electrical and Electronics Engineers, Piscataway NJ USA, pp. 149-158, International Conference on Software Engineering 2015, Florence, Italy, 16/05/15. https://doi.org/10.1109/ICSE.2015.143

Performance analysis using Subsuming Methods: an industrial case study. / Maplesden, David; Randow, Karl Von; Tempero, Ewan et al.
Proceedings - 2015 IEEE/ACM 37th IEEE International Conference on Software Engineering, ICSE 2015, Volume 2. ed. / Sebastian Elbaum; Gerardo Canfora. Vol. 2 Piscataway NJ USA: IEEE, Institute of Electrical and Electronics Engineers, 2015. p. 149-158 7202959.

Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review

TY - GEN

T1 - Performance analysis using Subsuming Methods

T2 - International Conference on Software Engineering 2015

AU - Maplesden, David

AU - Randow, Karl Von

AU - Tempero, Ewan

AU - Hosking, John

AU - Grundy, John

N1 - Conference code: 37th

PY - 2015

Y1 - 2015

N2 - Large-scale object-oriented applications consist of tens of thousands of methods and exhibit highly complex runtime behaviour that is difficult to analyse for performance. Typical performance analysis approaches that aggregate performance measures in a method-centric manner result in thinly distributed costs and few easily identifiable optimisation opportunities. Subsuming methods analysis is a new approach that aggregates performance costs across repeated patterns of method calls that occur in the application's runtime behaviour. This allows automatic identification of patterns that are expensive and represent practical optimisation opportunities. To evaluate the practicality of this analysis with a real world large-scale object-oriented application we completed a case study with the developers of letterboxd.com - a social network website for movie goers. Using the results of the analysis we were able to rapidly implement changes resulting in a 54.8% reduction in CPU load and an 49.6% reduction in average response time.

AB - Large-scale object-oriented applications consist of tens of thousands of methods and exhibit highly complex runtime behaviour that is difficult to analyse for performance. Typical performance analysis approaches that aggregate performance measures in a method-centric manner result in thinly distributed costs and few easily identifiable optimisation opportunities. Subsuming methods analysis is a new approach that aggregates performance costs across repeated patterns of method calls that occur in the application's runtime behaviour. This allows automatic identification of patterns that are expensive and represent practical optimisation opportunities. To evaluate the practicality of this analysis with a real world large-scale object-oriented application we completed a case study with the developers of letterboxd.com - a social network website for movie goers. Using the results of the analysis we were able to rapidly implement changes resulting in a 54.8% reduction in CPU load and an 49.6% reduction in average response time.

KW - Object oriented software

KW - Performance analysis

KW - Runtime bloat

KW - Subsuming methods

UR - http://www.scopus.com/inward/record.url?scp=84951774525&partnerID=8YFLogxK

U2 - 10.1109/ICSE.2015.143

DO - 10.1109/ICSE.2015.143

M3 - Conference Paper

AN - SCOPUS:84951774525

VL - 2

SP - 149

EP - 158

BT - Proceedings - 2015 IEEE/ACM 37th IEEE International Conference on Software Engineering, ICSE 2015, Volume 2

A2 - Elbaum, Sebastian

A2 - Canfora, Gerardo

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - Piscataway NJ USA

Y2 - 16 May 2015 through 24 May 2015

ER -

Maplesden D, Randow KV, Tempero E, Hosking J, Grundy J. Performance analysis using Subsuming Methods: an industrial case study. In Elbaum S, Canfora G, editors, Proceedings - 2015 IEEE/ACM 37th IEEE International Conference on Software Engineering, ICSE 2015, Volume 2. Vol. 2. Piscataway NJ USA: IEEE, Institute of Electrical and Electronics Engineers. 2015. p. 149-158. 7202959 doi: 10.1109/ICSE.2015.143

Performance analysis using Subsuming Methods: an industrial case study

Abstract

Conference

Keywords

Access to Document

Other files and links

Cite this