论文信息 - Using trace analysis for improving performance in COTS systems

Using trace analysis for improving performance in COTS systems

With Commercial off-the-shelf (COTS) based systems, developers are focused on "glue" code for integrating all the components to create applications. Existing tools for analyzing performance are not sufficient anymore with large systems. This paper describes new methods and tools for improving performance in COTS-based systems by analyzing the execution trace. The results of the analysis help the developer to tune his/her application and make the best usage of the underlying COTS components. A visualization tool integrates the analysis results with Eclipse, a major Java open-source IDE.

Erik Putrycz | E. Putrycz

[1] Fazel Famili,et al. Data Mining: Understanding Data and Disease Modeling , 2003, Applied Informatics.

[2] Clemens A. Szyperski,et al. Component software - beyond object-oriented programming , 2002 .

[3] Scott McFarling,et al. Reality-based optimization , 2003, International Symposium on Code Generation and Optimization, 2003. CGO 2003..

[4] J. Ross Quinlan,et al. C4.5: Programs for Machine Learning , 1992 .

[5] Rastislav Bodík,et al. An efficient profile-analysis framework for data-layout optimizations , 2002, POPL '02.

[6] Jack W. Davidson,et al. Profile guided code positioning , 1990, SIGP.

[7] B. Craig Meyers,et al. Managing Software Acquisition: Open Systems and COTS Products , 2001 .

[8] Surajit Chaudhuri,et al. Materialized view and index selection tool for Microsoft SQL server 2000 , 2001, SIGMOD '01.

[9] Douglas C. Schmidt,et al. Patterns for concurrent and networked objects , 2000 .

[10] Richard Johnson,et al. Proceedings of the international symposium on Code generation and optimization: feedback-directed and runtime optimization , 2003 .

[11] William G. Griswold,et al. An Overview of AspectJ , 2001, ECOOP.