• DocumentCode
    2385059
  • Title

    Component Monitoring of OSGi-Based Software

  • Author

    Wang, Tao ; Zhou, Xiaowei ; Wei, Jun ; Zhang, Wenbo ; Zhu, Xin

  • Author_Institution
    Technol. Center of Software Eng., Chinese Acad. of Sci., Beijing, China
  • fYear
    2010
  • fDate
    10-12 Nov. 2010
  • Firstpage
    250
  • Lastpage
    255
  • Abstract
    OSGi is a widely used service-oriented platform which provides support for the development and execution of component-based e-business applications. The component monitoring of OSGi-based software is important to the software reliability analysis and abnormality diagnosis. However, current approaches not only bring significant overhead, but also lack interaction behavior analysis. This paper focuses on monitoring components from resource utilization and interaction behavior perspectives. We propose an approach to trace thread execution for updating the relationship between threads and bundles, so that the CPU and memory utilization of bundles are calculated, and the service invocation graph is updated at runtime. A prototype tool is implemented and applied in an application server OnceAS based on OSGi. The experiments using TPC-W benchmark validate our approach without significant resource and performance overhead. Furthermore, simulation results show that our approach is of high accuracy for monitoring CPU utilization of bundles.
  • Keywords
    business data processing; service-oriented architecture; software reliability; CPU bundle utilization; OSGi based software; TPC-W benchmark; application server OnceAS; component based e-business applications; service oriented platform; software reliability analysis; Accuracy; Instruction sets; Java; Monitoring; Resource management; Servers; OSGi; monitoring system; resource utilization; service dependency; software component;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    e-Business Engineering (ICEBE), 2010 IEEE 7th International Conference on
  • Conference_Location
    Shanghai
  • Print_ISBN
    978-1-4244-8386-0
  • Electronic_ISBN
    978-0-7695-4227-0
  • Type

    conf

  • DOI
    10.1109/ICEBE.2010.15
  • Filename
    5704324