• DocumentCode
    2667232
  • Title

    Diagnosing Web Services System Based on Execution Traces Pattern Analysis

  • Author

    Mao, Chengying ; Tervonen, Ilkka ; Chen, Jinfu

  • Author_Institution
    Sch. of Software & Commun. Eng., Jiangxi Univ. of Finance & Econ., Nanchang, China
  • fYear
    2011
  • fDate
    19-21 Oct. 2011
  • Firstpage
    207
  • Lastpage
    214
  • Abstract
    Web service has been widely accepted as a technical standard for constructing software system which supports interoperable machine-to-machine interaction over a network. Even though such new system exhibits a lot of advantages such as loose-coupling, heterogeneity and flexibility, how to maintain it is still an open problem, especially for fault diagnosis. In the paper, a framework of debugging Web services system through intelligent information process methods is proposed. In such framework, the execution traces can be collected via SOAP message monitoring or service invocation instrument. Meanwhile, execution logic between Web services can be analyzed according to time-stamp. Based on the logic relations, service order matrix for each execution trace can be constructed. Then, cluster analysis technique can be performed on such matrices to partition the similar traces together. Subsequently, the failure reference 2-tuple which is the important clue for further fault location is formed through sampling and similarity analysis. Finally, the case study and simulation experiment are performed to confirm the feasibility and effectiveness of our method.
  • Keywords
    Web services; matrix algebra; open systems; pattern clustering; program debugging; program diagnostics; software fault tolerance; SOAP message monitoring; Web services system debugging; Web services system diagnosis; cluster analysis technique; execution logic; execution traces pattern analysis; failure reference 2-tuple; fault diagnosis; fault location; flexibility; heterogeneity; intelligent information process methods; interoperable machine-to-machine interaction; loose-coupling; sampling analysis; service invocation instrument; service order matrix; similarity analysis; software system construction; Business; Debugging; Educational institutions; Fault diagnosis; Simple object access protocol; Web services; clustering analysis; execution logic; execution trace; fault diagnosis; service order matrix;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    e-Business Engineering (ICEBE), 2011 IEEE 8th International Conference on
  • Conference_Location
    Beijing
  • Print_ISBN
    978-1-4577-1404-7
  • Type

    conf

  • DOI
    10.1109/ICEBE.2011.43
  • Filename
    6104619