<  Back to the Polytechnique Montréal portal

Diagnosing performance variations by comparing multi-level execution traces

François Doray and Michel R. Dagenais

Article (2016)

[img]
Preview
Accepted Version
Terms of Use: All rights reserved.
Download (498kB)
Cite this document: Doray, F. & Dagenais, M. R. (2016). Diagnosing performance variations by comparing multi-level execution traces. IEEE Transactions on Parallel and Distributed Systems, 28(2), p. 462-474. doi:10.1109/tpds.2016.2567390
Show abstract Hide abstract

Abstract

Tracing allows the analysis of task interactions with each other and with the operating system. Locating performance problems in a trace is not trivial because of their large size. Furthermore, deep knowledge of all components of the observed system is required to decide whether observed behavior is normal. We introduce TraceCompare, a framework that automatically identifies differences between groups of executions of the same task at the user space and kernel levels. Many performance problems manifest themselves as variations that are easily identified by our framework. Our comparison algorithm takes into account all threads that affect the completion time of analyzed executions. Differences are correlated with application code to facilitate the correction of identified problems. Performance characteristics of task executions are represented by a new data structure called enhanced calling context tree (ECCT). We demonstrate the efficiency of our approach by presenting four case studies in which TraceCompare was used to uncover serious performance problems in enterprise and open source applications, without any prior knowledge of their codebase. We also show that the overhead of our tracing solution is between 0.2 and 9 percent depending on the type of application.

Uncontrolled Keywords

Performance analysis, tracing, software visualization,concurrency, operating systems

Open Access document in PolyPublie
Subjects: 2700 Technologie de l'information > 2706 Génie logiciel
2700 Technologie de l'information > 2720 Logiciel de systèmes informatiques
Department: Département de génie informatique et génie logiciel
Research Center: Non applicable
Funders: CRSNG/NSERC
Grant number: CRDPJ468687-14
Date Deposited: 01 Feb 2018 11:00
Last Modified: 24 Oct 2018 16:12
PolyPublie URL: https://publications.polymtl.ca/2961/
Document issued by the official publisher
Journal Title: IEEE Transactions on Parallel and Distributed Systems (vol. 28, no. 2)
Publisher: IEEE
Official URL: https://doi.org/10.1109/tpds.2016.2567390

Statistics

Total downloads

Downloads per month in the last year

Origin of downloads

Dimensions

Repository Staff Only