ERA Banner
Download Add to Cart Share
More Like This
  • http://hdl.handle.net/10402/era.25346
  • Detecting And Diagnosing Web Application Performance Degradation In Real-Time At The Method Call Level
  • Wang, Mengliao
  • English
  • web application
    performance degradation detection
  • Jan 31, 2012 9:57 AM
  • Thesis
  • English
  • Adobe PDF
  • 2163421 bytes
  • As e-commerce becomes more popular, the performance of enterprise web applications becomes an important and challenging issue. Unlike failures, performance degradation is more difficult for the administrator to observe. It also takes much time to locate the root cause because there are many possible performance bottlenecks including network I/O, resource starvation, source code mistakes and even high site requests. In this thesis, we propose a system which is able to detect and diagnose web application performance degradations at the method call level in real time. We also set four sensors on the web server to monitor web application conditions, and use the measurements as input in detection and diagnostic algorithms. We implement our system in Java and build a test bed using a pet store application. Our system is evaluated on the test bed and results are encouraging, showing that our approach outperforms a traditional detection approach.
  • Master's
  • Master of Science
  • Department of Computing Science
  • Spring 2012
  • Wong, Kenny (Computing Science)
  • Bulitko, Vadim (Computing Science)
    Hoover, James (Computing Science)
    Musilek, Petr(Electrical and Computer Engineering)
    Wong, Kenny (Computing Science)