BI-OBJECTIVE FAULT TOLERANT MODEL FOR WORKFLOW TASK SCHEDULING ON GRIDS Sridevi S 1 , Golda Jeyasheeli P 2 1 PG student, Dept of CSE, Mepco Schlenk Engineering College, Sivakasi, India. 2 Assistant Professor, Dept of CSE, Mepco Schlenk Engineering College, Sivakasi, India, Email:L sridevi5983 @ gmail.com Abstract The spur of Grid computing is to aggregate the power of widely dispersed resources, and provide non-trivial services to users. In attempts to utilize a diverse set of resources in grids proficiently, scheduling has been made. The primary intention of scheduling is the minimization of application completion time; however, they may lead to the usage of excess and redundant resources. Our algorithm performs the scheduling by accounting for both completion time and resource usage. Since the performance of grid resources changes dynamically and the accurate estimation of their performance is very difficult, our algorithm incorporates rescheduling to deal with unforeseen performance fluctuations effectively. Also, fault tolerance is an essential part of the grid. In Grid environments, execution failures can occur for various reasons such as network breakdown, failure or non-availability of required resources. Fault tolerance can be achieved in grids by Over provisioning and Check pointing techniques. Since, over provisioning violates the resource usage control, check pointing strategy is implemented in our proposed method. Key Words: — check pointing, fault tolerance, grid scheduling, make span, resource optimization I. INTRODUCTION COMPUTATIONAL grid is a hardware and software infrastructure that provides dependable, consistent, pervasive, and inexpensive access to high-end computational capabilities. Recent developments in grid infrastructure technologies make it possible to execute large and distributed applications on it. Many of these applications fall in the category of interdependent task model. These classes of applications are generally referred as workflow applications, which are often represented as DAGs (Directed Acyclic Graph) with nodes representing tasks and edges representing dependencies. Scheduling algorithms in grid platforms generally focus on the minimization of application completion time. However resource usage also plays an essential role in the performance of grid. Scheduling on minimum make span with minimal use of resources yields optimal solution. To achieve non-trivial services from grids, an efficient fault tolerant strategy becomes vital. Hence our paper focuses on all these three aspects: Make span, resource usage, fault tolerance effectively. In this paper, we address the problem of scheduling workflow applications in grids and propose by a novel semi-dynamic scheduling heuristic, referred as adaptive bi-intentional fault tolerant scheduling (ABIFTS) algorithm. ABIFTS algorithm staticallygenerates the initial schedule using an evolutionary technique. It adapts dynamically as the performance of resources changes. Dynamic rescheduling also occurs where there is any fault in the schedulingprocess. ABIFTS can generate an optimized schedule even in presence of resource failure. The main strengths of the proposed system are: 1) The good quality of output schedules with minimum resource usage. 2) The adaptability to resource performance fluctuations. 3) The ability to generate efficient schedule in presence of resource failure. The first strength is achieved by iteratively improving an initial random schedule using a branch-and-bound technique and mutation. The second strength is made possible by using a rescheduling strategy. Specifically, ABIFTS initiates a rescheduling event if a job finishes later than expected and its late completion results in an increase in the overall application completion time. Rescheduling occurs with all of the remainingjobs that are not running. The third strength is achieved by means of check pointing strategy. Check pointing is a technique for inserting fault tolerance into computing systems. It consists of storing a snapshot of the current application state and uses it for restarting the execution in case of failures. International Journal on Information Sciences and Computing, Vol. 4, No.2, July 2010 43