Research and implementation of software behavior monitoring for J2EE |
|
Author | ShenLiXia |
Tutor | YanHan |
School | Nanjing University of Technology and Engineering |
Course | Applied Computer Technology |
Keywords | J2EE Probe Request path Component dependency graph Behavioral observation Figure |
CLC | TP311.52 |
Type | Master's thesis |
Year | 2011 |
Downloads | 15 |
Quotes | 1 |
Component technology development it is widely used in the development of enterprise-class applications , but the complexity of the structure of the enterprise-class component software component internal structure and system runtime behavior is difficult to fully grasp the cause developers and maintainers , greatly increase the debugging and maintenance difficulty. By the the monitoring component software run - time behavior , can accurately understand the interaction between components and dependencies , so crucial to the monitoring of component-based software . The monitoring function is achieved by modifying the source code of the target system flexibility and reusability , the paper proposes an applicable J2EE typical component-based software runtime behavior monitoring system RBMS (Runtime Behavior Monitoring System), does not depend on the monitoring system source code , high degree of modularity , reusability , flexible use . The RBMS through collection, analysis of the interaction between the internal components of the target system is running data to build the system run - time behavior of the model - component dependency graph , and behavioral observations in Fig . First identify the user requests by intercepting filter mode and assigns a unique identification for each request . Secondly, respectively, to achieve a non-distributed and distributed environment , the user request the implantable monitoring probe to intercept component interaction data , stored in the log in layers of tracking system to the target system . Finally, the analysis of the log , component interaction data path to request identity associated with the same request to build the system runtime behavior model , and presents visual results . The behavior of the model can be applied to many aspects , such as may be used for fault localization and detection of anti-patterns , has broad application prospects . J2EE Typical applications JPetstore expand experimental , small the experimental proof RBMS performance overhead , and can be widely used in the actual environment .