Reliability Based Workflow Scheduling on Cloud Computing with Deadline Constraint

Distributed computing workflow is an effective paradigm to express a range of applications with cloud computing platforms for scientific research explorations. One of the most difficult application areas of cloud computing technology is task scheduling. In a cloud, heterogeneous context, job scheduling with minimal execution cost and time, as well as workflow reliability, are critical. While working in the heterogeneous cloud environment, tasks that are successfully executed are widely identified by considering the failure of the processor or any communication technologies link. It will also have an impact on the workflow's reliability as well as the user's service quality expectations. This research paper proposes a Critical Parent Reliability-based Scheduling (CPRS) method that uses the reliability parameter to plan the task while taking into account the user-defined cost and deadline metrics. The effectiveness of the algorithm is compared to current algorithms utilizing scientific workflows as a benchmark, such as Cybershake, Sipht, and Montage. The simulation results supported the assertions by efficiently allocating resources to the cloudlets and stabilizing all of the aforementioned parameters using sufficient performance metrics growth.