Versatile provisioning and workflow scheduling in WaaS under cost and deadline constraints for cloud computing
Pon. Partheeban, V. Kavitha · Transactions on Emerging Telecommunications Technologies · 2018
Abstract Cloud computing environment delivers resources and services via a pay‐as‐you‐go billing model, and scheduling workflows have become a major application for distributed systems. An efficient scheduling is very challenging in a cloud to achieve minimum overhead, low cost, and high utilization of resources within the deadline. Workflow as a service (WaaS) provides a simple cost successive way to use the workflow applications in the cloud. This paper introduces a Gang of Tasks (GoTs)–based dynamic scheduling technique for scheduling the workflow to reduce the cost of renting virtual machines (VMs) under the deadline constraints in the WaaS environment. First, the GoTs strategy groups the workflows based on the priority and dependency of data. It schedules similar tasks into the same gang to execute the task from the gang simultaneously to the VMs. After scheduling the task in the GoTs, the deadline is distributed among all the gangs in the workflow. Then, the versatile resources are provisioned, dynamically depending upon the usage of containers in the VMs. The simulation results validate the ability to satisfy the deadline constraints and cost deficiencies of the proposed method as compared to the state‐of‐the‐art methods.