Competent scheduling is vital in exploiting the available computing resources. Scheduling dependant tasks is a well-studied problem and a number of approaches have been proposed. The problem of scheduling has been shown to be NP-complete. The existing algorithms for heterogeneous domains lag in efficiency as they do not consider the failures of the real world processors .The goal of the scheduling algorithm is to obtain a schedule with minimum makespan and maximum reliability. In this paper, two algorithms namely, Reliable Performance Effective Task Scheduling (RPETS) algorithm and Reliable Heterogeneous Earliest Finish-Time (RHEFT) algorithm, which are modifications of Performance Effective Task Scheduling(PETS) and Heterogeneous Earliest Finish-Time (HEFT) algorithms by including a reliability factor in the processor selection phase are proposed .The complexity of both the proposed algorithms is O(v2*p)
展开▼
