摘要
在面向服务的网格中,需要按用户请求动态地根据当前系统中服务的状况,组合出满足用户请求的服务组合。目前,当存在多个服务提供者提供相同功能的服务时,如何动态选择合适的服务却很少被论述,而且真实网格环境中的不确定性(如Cpu、网络的负载)也很少考虑。如何选择最合适的服务所在的资源,是服务组合中必须要考虑的。在分析动态服务组合的基础上,设计了一种动态选择资源的方法。并利用GT4中工厂设计模式给出其实现方式。该设计考虑了网格环境中资源的安全性和不确定性因素,比如资源经常失败,以及Cpu和网络的负载等,从而最大程度地提高服务执行的可靠性和高效性以及均衡系统负载。
In the service- oriented grid, need to compose basic services dynamically that have existed in the grid to form more sophisticated services according to the user's request. In the grid, however, multiple service providers can provide the same service with different implementation. Currently, dynamic service selection among similar services from multiple providers has seldom been addressed. And uncertainty(e, g. Cpu load and network load)under a real, dynamic grid environment has never been considered. How to choose the proper resottree associated with the required service is central to service composition. Based on the dynamic service composition, project a method of dynamic resource selection and provide it concrete implementation supported by factory/instance pattern in GT4. Considering the surety and uncertainty of the grid, the method can maximize the reliability and efficiency of service execution and balance the system work- load.
出处
《计算机技术与发展》
2007年第12期221-224,共4页
Computer Technology and Development
基金
江苏省高技术研究项目(BG2005020)
江苏省教育自然科学基金(04KKB320134)
关键词
网格
动态服务组合
资源选择
不确定性
工厂模式
grid
dynamic ,service composition
resource selection
uncertainty
factory instance