Workflow scheduling in cloud environment – Challenges, tools, limitations & methodologies: A review

Abstract

Workflow scheduling is a framework that simulates major domain applications like gaming, Natural language processing, data science etc., for which the task order automation and demand consistency to be maintained. This increases the complexities for CSPs based on various QoS parameters like cloud deployment models, service types, VM templates, migration process, energy dissipation etc., The recent research survey analysis and marketing perspective parameters include unknown demands, task requirements failure, cost incurs access delay, feasible deadlines, system capability, cache latency, scheduler types and policies, VMs, platform support, fault-tolerant and virtualization types and their limitations. The various survey papers considered were supporting QoS parameters of cloud environment towards industry task automation wherein the researchers will find their way to identify problem definitions with measured approaches and solutions. The analysis would pave way to incorporate machine learning approaches to derive workflows for business, healthcare, Speech recognition, Text recognition; drowsiness detection, road sign detection, metal part identification etc., The analysis done in this paper leads to various challenges and issues found in workflow scheduling approaches. The survey paper is summarized with introduction in chapter 1 followed by comparative conclusions of various research papers and concluded with summary of the findings based on the survey. This survey work can be extended considering each task in workflow and data migration types & issues which incurs budget, legal issues & policies and technical methodologies to find feasible solution for task automation.