针对服务网格环境中资源的动态性,提出了一种并行调度算法PGSWA(parallel grid service workflow scheduling),该算法引入了性能预测模型和并行就绪队列来预测下一段时间资源的性能并使得成员服务能够并行执行。实验证明,该算法能较好地缩短工作流的执行时间,提高工作流的执行性能。
Considering the dynamics of resources in the service-oriented grid environment, this paper proposed a novel scheduling algorithm PGSWA( parallel grid service workflow scheduling) to solve such problem. The algorithm introduced the performance prediction model and parallel ready queue into the scheduling to predict the resource performance and run services in parallel. The experiment results show that it can reduce the executing time of workflow, and improve its performance.