Survivability-based Scheduling Algorithm for Bag-of-Tasks Applications with Deadline Constraints on Grids

In the dynamic, complex and unbounded Grid systems, failures of Grid resources caused by malicious attacks and hardware failures are inevitable and can have an adverse effect on the execution of applications. Therefore it becomes an important and difficult issue to guarantee that the applications execute normally in Grid environment. To alleviate this problem, the survivability-driven scheduling algorithms are proposed in our previous work. However these algorithms don't consider the deadline requirements of users. In this paper, we propose a Survivability-Based scheduling algorithm for bag-of-tasks applications with Deadline Constraints that maximizes the survivability while meeting the deadline for delivering results, which is referred as SBDC algorithm. Compared with traditional scheduling algorithms, this algorithm is more adaptable to the complex Grid computing environment. Experimental results reveal that this algorithm can maximize the survivability of applications while meeting the application deadline.