Performance regression refers to the phenomena where the application performance degrades compared to prior releases. Performance regressions are unwanted side-effects caused by changes to application or its execution environment. Previous research shows that most problems experienced by customers in the field are related to application performance. To reduce the likelihood of performance regressions slipping into production, software vendors must verify the performance of an application before its release. The current practice of performance verification is carried out only at the implementation level through performance tests. In a performance test, service requests with intensity similar to the production environment are pushed to the applications under test; various performance counters (e.g., CPU utilization) are recorded. Analysis of the results of performance verification is both time-consuming and error-prone due to the large volume of collected data, the absence of formal objectives and the subjectivity of performance analysts. Furthermore, since performance verification is done just before release, evaluation of high impact design changes is delayed until the end of the development lifecycle. In this thesis, we seek to improve the effectiveness of performance verification. First, we propose an approach to construct layered simulation models to support performance verification at the design level. Performance analysts can leverage our layered simulation models to evaluate the impact of a proposed design change before any development effort is committed. Second, we present an automated approach to detect performance regressions from results of performance tests conducted on the implementation of an application. Our approach compares the results of new tests against counter correlations extracted from performance testing repositories. Finally, we refine our automated analysis approach with ensemble-learning algorithms to evaluate performance tests conducted in heterogeneous software and hardware environments. / Thesis (Master, Electrical & Computer Engineering) -- Queen's University, 2011-01-31 15:53:02.732

Doctor en Ciencias, Mención Computación / Los cambios continuos en el código fuente de un programa pueden inadvertidamente introducir una regresión de rendimiento en tiempo de ejecución. Dichas regresiones se refieren a situaciones donde el rendimiento de un programa se degrada en comparación de versiones anteriores del mismo, aunque la nueva versión funcione correctamente. Ejecutar puntos de referencia en cada versión de un programa es una técnica tradicional utilizada para identificar regresiones en etapas tempranas. A pesar de ser efectiva, esta técnica exhaustiva es difícil de llevar a cabo en la práctica, principalmente por la alta sobrecarga que esta actividad demanda. En esta tesis, realizamos un estudio empírico sobre una variedad de programas, con el fin de evaluar cómo el rendimiento de un programa evoluciona en el tiempo, a medida que es modificado. Guiados por este estudio, proponemos Horizontal Profiling, una técnica de muestreo para inferir si una nueva versión de un programa introduce una variación de rendimiento, usando información de la ejecución de versiones anteriores. El objetivo de Horizontal Profiling es reducir la sobrecarga que requiere monitorear el rendimiento de cada versión, ejecutando los puntos de referencia solo en las versiones que contengan cambios costosos de código fuente. Presentamos una evaluación en la cual aplicamos Horizontal Profiling para identificar regresiones de rendimiento en un número de programas escritos en en el lenguaje de programación Pharo. En base a las aplicaciones que analizamos, encontramos que Horizontal Profiling es capaz de detectar más del 80% de las regresiones de rendimiento, ejecutando los puntos de referencia en menos del 20% de las versiones. Adicionalmente, describimos los patrones identificados durante nuestro estudio empírico, y detallamos cómo abordamos los numerosos desafíos que enfrentamos para completar nuestros experimentos. / Este trabajo ha sido parcialmente financiado por CONICYT a través de la beca CONICYT-PCHA/Doctorado Nacional para extranjeros/2013-63130199, OBJECT PROFILE y LAM RESEARCH

Title: Influence of selected conditional factors on performance in white water slalom. Aims: The aim of the study was to investigate the relationship of selected specific movement abilities being examined modified test battery with the performance of athletes in the water slalom. Methods: It was used field measurements where the applied modified test battery. Using GPS module to determine the distance partial tests from batery. For measuring was used manual measurement. To determine the statistical correlation between the modified battery and performance ability of competitors was used two different coefficients of correlation and regression analysis. According to the order of the test and the race was used nonparametric correlation study - Spearman correlation coefficient. Determining the statistical significance of the relationship of measured times in tests and final time in the nomination races have used the Pearson correlation coefficient. Results: For a statistically significant relationship was determined value when r ≥ 0.8. Spearman's correlation coefficient: In the test at 40 m were obtained these correlation coefficients: Nomination races rs = 0,380952, Czech cup rs = 0,595238. In the test at 80 meters they were obtained these correlation coefficients: nomination races rs = 0,857143,...