Building an Intelligent System for Predicting and Fixing Performance Defects
Loading...
Authors
Zhao, Guoliang
Date
Type
thesis
Language
eng
Keyword
Performance anomalies , Performance defects , Pull request ranking , Fixing effort prediction
Alternative Title
Abstract
Software systems have been playing an essential role in supporting our daily activities. Performance anomalies are unexpected performance degradation that deviates from the normal behaviors of software systems. Performance anomalies can cause a dramatically negative impact on users' satisfaction. The increasing scale and complexity of software systems make these systems prone to performance anomalies that are caused by various reasons, such as, misconfiguration, hardware failures, resource contentions, and performance defects. To help development and operation teams to maintain the performance of software systems, prior studies propose various approaches to detect performance anomalies and performance defects. However, prior detection approaches cannot predict the performance anomalies ahead of time; such limitation causes an inevitable delay in taking corrective actions to prevent performance anomalies from happening.
To help developers and operators to prevent anomalies from happening, in this thesis, we conduct a set of studies to predict performance anomalies from run-time monitoring data and predict performance defects at the development phase. More specifically, our approach consists of four aspects: (1) we propose an approach to predict performance anomalies in software systems by analyzing run-time monitoring data; (2) we propose an approach that can predict a large variety of performance defects during development phase; (3) we provide a generic approach that predicts methods with any types of defects (e.g., performance and non-performance defects) and their fixing effort; and (4) we propose an approach to prioritize pull requests to help reviewers review code changes. Through a series of experiments, we observe that our approaches can help the development and operation teams to avoid performance anomalies at the run-time, and capture and fix performance defects at the development phase.
Description
Citation
Publisher
License
Queen's University's Thesis/Dissertation Non-Exclusive License for Deposit to QSpace and Library and Archives Canada
ProQuest PhD and Master's Theses International Dissemination Agreement
Intellectual Property Guidelines at Queen's University
Copying and Preserving Your Thesis
This publication is made available by the authority of the copyright owner solely for the purpose of private study and research and may not be copied or reproduced except as permitted by the copyright laws without written authority from the copyright owner.
ProQuest PhD and Master's Theses International Dissemination Agreement
Intellectual Property Guidelines at Queen's University
Copying and Preserving Your Thesis
This publication is made available by the authority of the copyright owner solely for the purpose of private study and research and may not be copied or reproduced except as permitted by the copyright laws without written authority from the copyright owner.