Research On Task Scheduling Methods Of Collaborative Design In Grid Environment

As a new technology used to provide high-performance computing, management, and service, Grid Computing has attracted more and more attention. Furthermore, scheduling is a key factor that acts on the efficiency of Grid computing, as well as one of the most fundamental and essential problems that challenge us most. In the Collaborative Designing in Grid environment, scheduling is also one of the most fundamental and essential problems, and because of the characteristic of Collaborative Designing, the scheduling has its particularity.Resource reservation is an effective way to improve the efficiency of scheduling because it will facilitate the execution of large tasks or tasks with QoS desire, it also ensure the tasks get their resource for execution. In the process of Collaborative Designing in Grid environment, the tasks always vary in size, and the execution of some large tasks becomes the sticking point, fortunately, resource reservation can improve this problem. In that, resource reservation is adopted in the mechanism of task scheduling of Collaborative Designing in Grid environment.In this paper, aiming at the fact that some large subtasks in the DAG scheduling could delay the completion time of application task, the definitions of key task and resource-reserving opportunity along with the idea of dynamically reserving resources for key tasks was brought forward, which could actually reduce the completion time of application task. In this paper, on the basis of dynamical resource reservation, a task-mapping algorithm based on DAG in Grid environment named KtRR is provided. The results of experimentations show that the KtRR is more efficient than the existing algorithms and is able to be applied in task scheduling of Collaborative Designing in Grid environment.In this paper, according to the situation that most research on resource reservation focus on the computational resource, a unified resource reservation policy is provided to consider the reservation of other resources with computational ones at one time, by the virtue of this policy, the efficiency of reservation is enhanced, as the result, the makespan of task scheduling and is shorten. Simulation results shows, thepolicy works well with the algorithm mentioned previously which based on resource reservation, and the policy can be applied to the dynamic and heterogeneous resource environment of Grids.