Energy and Reliability aware Cloud Workflow Scheduling with Uncertain Task Execution Times

Energy consumption and reliability are crucial for cloud workflow scheduling, and many methods have been proposed to address these issues. However, these methods typically assume that task execution times are deterministic, and the problem of reliably scheduling workflows with uncertain task execution times while minimizing energy consumption has not yet been studied. To address this gap, this paper proposes a new algorithm named EWSR, which consists of three phases: workflow analysis, task scheduling, and online adjustment. The algorithm ensures the reliability of cloud workflows under uncertain task execution times while simultaneously reducing energy consumption. An online adjustment mechanism is designed to dynamically adjust task reliability and release idle resources, thereby increasing resource utilization while lowering energy consumption. Experimental results demonstrate that EWSR effectively reduces energy consumption while meeting workflow reliability requirements.