Workflow Scheduling Based on Deadline Constraints in Cloud Environment

loud computing is providing an environment for scientific workflows where large-scale and complex scientific analysis can be scheduled onto a heterogeneous collection of computational and storage resources. A scientific workflow is described as a paradigm, which is used to describe a set of structured activities and scientific computations. Scientific workflow scheduling has become one of the most challenging issues in cloud systems. Scheduling of scientific workflow applications involves the mapping of tasks to computational resources, based on quality of service requirements such as time, cost, bandwidth, etc. Most of the proposed scheduling algorithms require detailed information about tasks, e.g., execution time, and remaining time. On the other hand, the most proposed algorithms cannot schedule tasks in the shortest possible time by using minimum knowledge about tasks. In this article, we introduce an approach for task scheduling, namely RRRSD (Relation aware Round Robin Scheduling based on Deadline constraints). It applies the Round Robin algorithm along with deadline parameters. The main goal of this model is to optimize the mapping of tasks to available resources in order to minimize makespan time and the failure rate of scientific workflows. The simulation results show an average improvement of 24.25% for makespan time of workflows and the failure rate of 36.21% compared to four basic scheduling algorithms.